(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Biodiversity Heritage Library | Children's Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Basic Infrared Spectroscopy"

i 

o" 



a 

w 
■o 
to 
o 



1/5 

n 
o 

■D 



Basic Infrared 
Spectroscopy 



SECOND EDITION 



J.H.vanderMaas 



HEYDEN 





I 



fV 




V 



\_L- 



Basic Infrared 
Spectroscopy 



by J. H. van der Maas 

Analytisch Chemisch Laboratorium der / >! 
Rijkuniversiteit Utrecht > 

(*l^sl V : -.J .-#. 



SECOND EDITION 



HEYDEN B SON LTD 



London • New York • Rheine 



Heyden & Son Ltd., Spectrum House, Alderton Crescent, London NW 3XX 
Heyden & Son Inc., 225 Park Avenue, New York, N.Y. 10017, U.S.A. 
Heyden & Son GmbH, Steinfurter Strasse 45, 4440 Rheine/Westf., Germany 



© Heyden & Son Ltd., 1969, 1972. 
All Rights Reserved. No part of this publication 
may be reproduced, stored in a retrieval system, 
or transmitted, in any form or by any means, 
electronic, mechanical, photocopying, recording 
or otherwise, without the prior permission of 
Heyden & Son Ltd. 



Library of Congress Catalog Card No. 70-101090 
ISBN 85501 029 (paperback) 
ISBN 85501 031 2 (cloth) 



Made and printed in Great Britain at The Pitman Press, Bath 





Contents 






Foreword 


by G. Dijkstra 




vi 


Preface 






vii 


Chapter 1 


Introduction 




1 


Chapter 2 


Theory 




6 


Chapter 3 


Spectrophotometers 




31 


Chapter 4 


Sampling 




48 


Chapter 5 


Interpretation of Spectra 




65 


Appendix A Reference Spectra 




68 


Appendix B Typical Band Contours 




72 


Appendix C Correlation Charts 




82 


Appendix D Infrared Absorption Frequencies 
Groups 


of Functional 


91 


Appendix 


E General Bibliography 




105 


Index 






106 



Foreword 



Infrared spectroscopy is one of the branches of the physical sciences where practice 
ran far and wide, leaving theory struggling behind. Making sense of the spectra of 
large molecules involves theoretical methods often different from those which allow 
satisfactory treatment of the spectra of small molecules, and I believe that in spite 
of the venerable age of analytical infrared spectroscopy (a quarter of a century) new 
fruitful approaches will be found. This book, though not breaking new ground in 
infrared science, is an attempt to bring an understanding of the methods one uses in a 
reasoned approach to the interpretation of the spectra of large molecules to the 
novice in the field of analytical infrared spectroscopy; the author has attempted to 
restrict the subject matter to the necessary elements of this fascinating branch of 
science, but to give these a full treatment. May the quality of the work done by those 
who have used this book be his reward. 

G. Dijkstra 



Preface 



The second edition of this book is only slightly different from the first one. Suggestions 
and critical remarks have been taken into account, while on request a general biblio- 
graphy has been added containing several monographs devoted to different applica- 
tions of infrared spectroscopy. 

Appendix B has been enlarged with a few more typical band contours and explana- 
tory text has been added. 

With respect to the correlation tables a few remarks may be necessary. In general, 
present-day tables and charts fail for three reasons mainly : 

(i) all published data are used without attention to the difference in accuracy. 
The regions therefore become broader and broader, while simultaneously 
the fine structure in a certain group gets lost. 

(ii) in preparing the tables no allowance is made for the various sampling tech- 
niques, though it is well known that significant frequency shifts occur in 
different solvents. 

(iii) sometimes data from the literature quoted are inaccurate or false, probably 
initially by accident but persisting thereafter due to carelessness. 

The remedy seems to be simple but laborious: new charts and tables are to be 
prepared; e.g. tables for gases, for pure liquids, for solids in KBr, etc. The accuracy 
should be known. We believe this to be a necessary operation for there is a real 
danger that infrared spectroscopy will become ineffective as a tool for structure deter- 
mination. Tables as such are in preparation but not yet ready. For the time being the 
less invaluable ones are to be used. 

I would like to express my gratitude to Dr. M. A. Ford of Perkin-Elmer, Beacons- 
field, for his interest and stimulating remarks and to Mr. E. T. G. Lutz for his help in 
preparing the additional spectra for this edition. 

The co-operation of the publisher is gratefully acknowledged. As before, remarks 
and suggestions to improve the usefulness of this book will be very much appreciated. 

The Author 



Units and abbreviations 



Variable 



Symbol 



Unit 



Unit Abbreviation 



length 
time 


/ 
t 


mass 


m 


temperature 

energy 

intensity 


T 
E 
I 


frequency 

wavelength 

wavenumber 


V 

A 

a 


index of refraction 


n 


transmission 


T 



metre 


m 




secona 


s 




gram 
degree 
erg or joule 
joules per 
second 


g 
Kor 

erg or 

Js" 1 


J 


hertz 


Hz 




micrometre 


(im 




reciprocal 
centimetres 


cm -1 





percent transmission %T 



Prefixes 



Prefix 


Abbreviation 


Value 


mega 


M 


10 6 


kilo 


k 


10 3 


milli 


m 


IO- 3 


micro 


V- 


io- 6 


nano 


n 


10- 9 


pico 


P 


10- 12 



Conversion Factors and Constants 

light velocity in vacuo (c) = 3 x 10 8 m s - 1 
^ lnet £ corresponds to ^ hoton m 



100 

300 

1000 



144 
48 
14-4 



^photon at 1 A* m = 28-7 kcal/mole 
^photon at 1 nm == 2-87 x 10 7 cal/mole 

or = 1-99 x 10" 16 J/mole 

1 A (angstrom) = 100 pm = 0-3 x 10 19 Hz 




1 

Introduction 



ELECTROMAGNETIC RADIATION 

Electromagnetic radiation extends from y- and cosmic rays to radio frequencies and 
includes the ultraviolet, visible and infrared regions. It has been proved to be a periodi- 
cally changing or oscillating electric field in a certain direction with a magnetic field 
oscillating at the same frequency but perpendicular to the electric field. One cannot 
produce one without the other (Fig. 1). The magnetic field will not concern us here. 




Fig. 1. Electromagnetic wave. H magnetic field, E electric field, A wavelength 



Electromagnetic radiation may be considered as a wave motion or as a stream of 
particles, often called quanta or photons. As a wave motion it can be characterised by 
a few parameters: the length of a wave, the wavelength (A); the speed at which the 
wave moves, i.e. the velocity (c); and the frequency (v) being the number of waves 
or cycles per second. As for sound waves, the velocity for electromagnetic waves proves 
to be a constant for the medium in which the waves are propagating (c = 3 X 10 8 
m s" x in vacuo). 



It can be seen easily that 



Xv = c 



the wavelength being inversely proportional to the frequency. Turning back to the 
corpuscular or quantum character of electromagnetic radiation we have from quan- 
tum mechanics the relation 



E=hv 
1 



2 ' INTRODUCTION 

(E = energy per quantum of radiation, h = Planck's constant). Thus v is related 
linearly to the energy of the radiation. 

The electromagnetic spectrum can be divided into several regions (Fig. 2) differing 
in frequency only, but each with its own special character. The static field corresponds 

Hz 

r- 

Alternating current 
t- 10 3 



- 10* 



- 10 y 



- 10 



12 



- 10 



,15 



- 10 



,18 



I— 10 



21 



Radio frequencies 



Micro waves 



Infrared radiation 



Visible light 
Ultraviolet light 



X- rays 



y-rays 



Nuclear Quadrupole Resonance 

Nuclear Magnetic Resonance 

Electron Spin Resonance 



Rotation 



Vibration 



Outer- electron transition 



Inner - electron transition 



Nuclear transition 



Fig. 2. Electromagnetic spectrum. The right hand side lists some possible spectroscopic 
sources and absorptions 



to v = 0. The highest frequencies are found at about 10 20 Hz. Infrared radiation is 
to be found at frequencies of 10 14 -10 12 Hz, corresponding to wavelengths of 1-100 fim 
or wavenumbers of 10 000-100 cm -1 . 

At room temperature (300 K) the average velocity of a gas molecule is about 500 
m/sec. So in 10 -7 s a molecule moves 50 jum. In the same time a radio wave with a 
frequency of 10 7 Hz completes one cycle. This means that the alterations in the electric 



field are followed easily by the relatively fast-moving molecules ; seen from the centre 
of gravity of the molecule the radio frequencies seem to be more or less 'static'. 
At frequencies of about 10 12 Hz, however, it is just the other way round: the molecule 
seems to be static with respect to the fast-oscillating electric field. Interaction of 
electromagnetic radiation and molecules will be possible in two ways : with or without 
energy exchange. 

REFRACTION AND DISPERSION 

Apart from the rotation of polarised light, refraction (or dispersion) is one of the few 
mechanisms of interaction between molecules and electromagnetic radiation without 
exchange of energy. It is the result of different velocities in different media. The index 
of refraction is given by 

velocity of light in vacuo 
velocity of light in a medium 

This index n is not a constant but depends to some extent on the wavelength of the 
radiation. In general n and d«/d^ decrease in magnitude as I increases (Fig. 3). It 



1-5 



1-4 



400 



500 



600 



700 



X(nm) 



Fig. 3. Refractive index of a non-absorbing medium as a function of the wavelength. The 
line can be represented by the Cauchy (1836) expression: n = A + B/X 2 + C/A 4 , where A, 
B and C are constants characteristic for the medium concerned 



is found, however, that as one approaches and passes wavelengths at which the medium 
absorbs, the index of diffraction as well as dn/dX are subject to great alterations (Fig. 
4). This is often referred to as anomalous dispersion. Unfortunately, one cannot make 
use of this high dispersive power because absorption increases too, and thus too much 
radiation will be absorbed if prisms of normal dimensions are used. It is worth noting, 
however, that the highest dispersion is found near to an absorption band at its 'long- 
wavelength side'. 

ABSORPTION 

Absorption is a form of interaction between matter and light or radiation in which 
energy exchange is involved. We will restrict ourself to the absorption of infrared 
radiation by molecules. Under what condition(s) can absorption take place ? 



INTRODUCTION 



As we know already, infrared radiation is a moving oscillating electric field, 
oscillating at frequencies of 10 12 -10 14 Hz. The internal movements of the atoms in a 
molecule with respect to its centre of mass occur at the same frequency. Let us con- 
sider a mechanical model first. Suppose one has a ball connected to the ceiling by an 



Absorption X 

bond 

Fig. 4. Anomalous dispersion in an absorption band 

elastic rubber string and suppose the ball is moving up and down. The ball will 
naturally come to rest sooner or later due to the gravitation force and the friction with 
the air. Now to keep it on moving one has to deliver energy to the system. This can be 
done by flipping the ball gently at the moment it starts moving upwards. In other 
words, one has to supply energy at the same frequency at which the system is moving. If 



1 m, 


















To 


r B +A 


r o 




r-A| 


r„ 


r c +A 


r o 




k 




„ i 


> 




i 


> „ 


fc i 


> 



one complete cycle in t s 



Fig. 5. Vibration of a diatomic molecule. m% = 2mi: z, centre of mass; ro, distance between 
mi and mi in the equilibrium position 



this is done, the maximum distance from ball to ceiling will steadily increase. If one 
just touches the ball or hits it at another frequency, the system will soon come to rest. 
Now consider a diatomic molecule with masses wi and m% at an equilibrium 
distance ro (Fig. 5). If such a molecule is vibrating, the masses wi and m% are moving 
towards and away from each other, i.e. the distance r is subject to periodic changes. 
The centre of mass of the molecule is considered to be fixed, otherwise translation would 
be involved. To keep this centre at one place, atom 1 must move twice as much as atom 2, 



due to the difference in mass. Assuming a difference in electric charge between mi and 
m 2 , a dipole exists for this molecule, directed for instance as shown in the diagram. 
During the vibration of the molecule the dipole moment will change simultaneously 
with the distance r. As this change is periodical, such a vibrating molecule produces a 
stationary alternating electric field, the frequency being 1/t Hz. As in the mechanical 
model this system can absorb energy provided it is supplied in a similar way, i.e. by 
a moving electromagnetic field oscillating at the same frequency 1/t as that of the 
molecule itself. 

If, however, this molecule has no alternating dipole, no oscillating electric field 
would arise and therefore no energy could be absorbed from any external source, 
in spite of the fact that this electromagnetic radiation might be of the correct frequency. 

SELECTION RULES 

The vibration of the diatomic molecule is a very simple one. For polyatomic mole- 
cules many vibrations exist, as we will see later. Nevertheless, what holds for the 
vibration of the diatomic molecule holds for any vibration in any molecule. This 
enables us to state some rules : 

1 . Absorption of infrared radiation by a vibrating molecule will only take place 
if the vibration produces an alternating electric field; i.e. if the vibration is 
coupled with a changing dipole moment. 

2. In order for the radiation to be absorbed, the vibration frequency of the mole- 
cule must be identical to the frequency of the radiation, and furthermore 
since E = hv, the absorbed quantum of energy will have a distinct value. In 
other words the absorption of energy is quantised. 

DISSIPATION OF ABSORBED ENERGY 

The energy absorbed by a molecule is rapidly dissipated. The excited molecule loses its 
energy (vibrational and rotational) in less than lO -6 s. The energy is either trans- 
formed into kinetic energy as result of collisions or released again as a photon. As the 
direction of the liberated photon is random in space, and because the absorption 
process can be repeated for such a photon on its way through the medium, it can be 
seen that for a once-absorbed photon the probability of re-emerging from the medium 
in the direction of the transmitted beam is negligibly small. 

BIBLIOGRAPHY 

W. Briigel, An Introduction to Infrared Spectroscopy, Methuen, London, 1962. 
R. P. Bauman, Absorption Spectroscopy, John Wiley and Sons, New York, 1962. 



2 
Theory 



Any movement of an atom in space can be represented by 3 independent mutually 
perpendicular movements parallel to the x, y and z axes in a cartesian system. The 
atom is said to have 3 degrees of freedom. A system of N free-moving atoms thus 
will have 37V degrees of freedom. 

Like an atom, the movement of any object in space, or rather the movement 
of its centre of mass, also called translation, can be described by 3 parameters. 
In addition, the object may rotate about the centre of mass. This rotation can also be 
represented by 3 parameters, i.e. three rotations on three cartesian axes originating 
in the centre of mass. 

It follows that to describe the movements of an object in space 6 degrees of freedom 
are required : three for translation and three for rotation. The same holds for a mole- 
cule of N atoms. For rotational and translational movements of the molecule 6 degrees 
of freedom are required, thus leaving 37V— 6 degrees for the movement of the atoms in 
the molecule, i.e. for the vibrations of the molecule. 

VIBRATION 

The movements of the atoms in a molecule may be very complex but as was shown 
above these complex vibrations can be composed from 37V— 6 basic vibrations, the 
so-called 'normal' or 'fundamental' vibrations. Provided these 37V— 6 vibrations 
satisfy the earlier stated selection rules, the molecule will give rise to at least 37V— 6 
absorption bands somewhere in the infrared region. For strictly linear molecules, 
such as carbon dioxide, the number of normal vibrations is 3N—5. Consider the three 
rotational axes of such a molecule. The position of one or more atoms will be changed 
if rotation about two of these axes is carried out. Rotation about the third axis, the one 
which coincides with the molecular axis, does not change the position of the atoms. 
Hence only 2 parameters - two degrees of freedom - are required to describe any 
rotation of such molecules, so leaving 3N—5 degrees of freedom for vibrational 
analysis. 

Diatomic molecules 

The number of normal or fundamental vibrations for this type of molecule is just 
one. According to the selection rules this vibration will only give rise to absorption 
of radiation in cases where the atoms are different, e.g. HC1, CO, IC1, NO etc. For 
only then can a dipole moment exist. The symmetric molecules such as H2, N2, O2, 
CI2 etc. will not absorb, as there will be no changing dipole moment. 

6 



Harmonic oscillator. At what frequency v will a diatomic molecule have its normal 
vibration? Consider the earlier-mentioned mechanical model: two masses mi and 
m% connected by a spring at a distance ro. If this distance has to be increased or 
decreased by Ar a force F has to be applied which is proportional to Ar (Hooke's 
law for a harmonic oscillator) : 

F=-/Ar 

where / is the proportionality factor or force constant. This holds as long as Ar is 
reasonably small. So if this system is oscillating, Ar being small, a simple harmonic 
motion will be the result. The frequency of such an oscillation is known to be given by : 






where ju, is the reduced mass, to be found from 



n = 



nti + ni2 



or 



i--L + -L 

f* mi mi 



As can be seen, the frequency of a harmonic oscillator depends only on the force 
constant /and the reduced mass // (see Table 1). 

Table \ a 



Molecule 


fi 


/ 


a(v) 


2B Q 




(1-66 x 10" 24 g) 


(10 5 dyne/cm) 
(Ncm- 1 ) 


(cm- 1 ) 


(cm- 1 ) 


H 2 


0-50 


507 


4160* 


61 


HD 


0-67 


515 


3631" 


46 


D 2 


100 


5-24 


2993 b 


30 


35 CI 2 


17-50 


3-21 


556 b 


0-3 


N 2 


700 


22-2 


233P 


2-0 


o 2 


800 


11-3 


1555 b 


1-4 


HF 


0-95 


8-62 


3935 c 


21 


H 35 C1 


0-97 


4-74 


2886 


10-6 


HBr 


0-98 


3-78 


2558 


8-5 


HI 


0-99 


2-89 


2233 


6-6 


NO 


7-46 


15-4 


1877 


1-7 


CO 


6-85 


18-6 


2143 


2 



a The data are taken from K. W. F. Kohlrausch, Der Smekal-Raman-Effect, and from G. Herzberg, 
Molecular Spectra and Molecular Structure. (See bibliography for this chapter.) 
b Taken from Raman spectroscopy as these vibrations are infrared inactive. 
c Extrapolated value. 

Energy. What is the total energy of such a classical harmonic oscillator? Suppose 
there will not be any friction losses due to any kind of interaction. Once moving, 
therefore, the system will maintain a constant total energy E, being the sum of the 
varying kinetic and potential energies, Tand V, i.e. E = T + V. According to Hooke's 
law the potential energy V is at any time : 

K=i/(Ar)2 



8 THEORY 

A curve representing V as a function of Ar will be a parabola (Fig. 6). Suppose the 
distance r between m\ and m% during the oscillation is varying from ro + Ar max to 
ro — Ar max , Ar max being the maximum displacement. When the distance is ro + Ar max 
or ro — Ar max the kinetic energy of the system is zero, there being no movement for 
an infinitely small moment; hence E = |/(Ar max ) 2 . 

When the distance mi — mi is ro the potential energy is zero and the kinetic 
energy T = £/(Ar max ) 2 . 

From the foregoing we may conclude that the total energy of a classical harmonic 
oscillator depends only on the force constant / and the maximum displacement 
Ar m a X . Hence any E value is allowed, for Ar max can take any value; E might therefore 
be thought to be continuously variable. 




-3 -2 -1 



Fig. 6. The potential energy (V) for a harmonic diatomic oscillator as a function of Ar 



From quantum-mechanical considerations however, one knows the energy 
values for a diatomic molecule are certainly not continuous. Equally spaced energy 

levels exist, as given by : 

E n = in + \)hv 

where n = 0, 1, 2, 3 . . . (any positive integer). Two adjacent levels differ by an 
energy hv, while the lowest level E is called the 'zero-point' energy. The quantum 
hv is in agreement with what is said in the selection rule, although it seems to be in 
contradiction with what has been found for the mechanical model, v for the mechanical 









n = ? F :khv 


' 


I i 




i 


1 


n = F :k hv 


o - 



Fig. 7. Energy levels for a diatomic harmonic oscillator 

model (masses in g) will be extremely small compared with v of the molecule (masses 
in 10 -23 g); in fact about 10 11 times smaller if only the masses are considered. Since 
the energy is proportional to v, the spacing between the energy levels in the case of the 
mechanical system will be so small that a continuum seems to be produced. 

A molecule in the ground state £b, absorbing an energy quantum hv, will reach 
the first excited state E\. The same molecule may absorb again a quantum hv and so 
reach the second excited state Ei. If, however, the molecule in the ground state ab- 
sorbed a quantum hv' , where v' = 2v it would have reached level E 2 at once (Fig. 7, 
dotted line). A condition from quantum mechanics states, however, that the only 
allowed transitions for a harmonic oscillator are those between adjacent levels or 

AE = En+i — E n = hv 



For a harmonic oscillator the transition 
is forbidden, while 



+ hv' 
Eq > E2 



+hv +hv _ 

Eo ► Ei > £2 



10 



THEORY 



is allowed. In the latter case both quanta have the same frequency and therefore 
the absorptions will occur at the same place in the infrared. Both absorptions will be 
indistinguishable. Furthermore, in order for a transition to occur between E\ and 
£2, a molecule which has just absorbed a quantum hv must meet a second quantum 
before the gained energy has been dissipated. Since the number of excited (£1) mole- 
cules is low according to Boltzmann's distribution law 



Nx 

}Vo = 6Xp \ kT 



( -(.Ei - Eo) \ 
I kT ) 



the probability for a transition £1 -> £2 will be low. N1/N0 is the ratio of molecules 
in level £1 to those in level £0, while A£ = hv = hco. At room temperature (~300 K) 
and for a = 900 cm -1 this ratio will be about 1 :77. For 600 and 1200 cm -1 this ratio 




Fig. 8. Energy levels for an anharmonic diatomic oscillator. 
parabola, potential energy curve 

is 1:18 and 1:300 respectively. Thus spectra of doubly excited molecules may be 
observed at low frequencies. 

Summarising the foregoing sections, we may conclude that a diatomic molecule 
acting as a harmonic oscillator appears to have but one transition. This is the basic 
or fundamental absorption, i.e. the one from the ground level £0 into the first excited 
state £1. In general these fundamentals fall between 4000 cm -1 and 400 cm -1 , the 
principal infrared region. 



11 

Anharmonic oscillator. In actual practice the vibration of the diatomic molecule 
is not strictly harmonic. The potential energy curve as a function of the displacement 
from the equilibrium distance ro is in fact not a parabola though it shows a close 
resemblance (Fig. 8). In the low-energy region the parabola is a good approximation 
to the true energy curve. The higher energy levels including the second are more closely 
approximated by 

E n = hv[(n + i) - x(n + $*] 

x being the anharmonicity constant, usually a small fraction of unity. The spacing 
between two levels is given by 

E n — En-i = AE = hv{\ — 2nx) n = 1, 2, 3 . . . 

and hence will differ from level to level. For the transition Eq — >• E\ the energy dif- 
ference will still be hv provided x is much smaller than unity. 

Overtones. As we have a harmonic oscillator no longer, transitions from Eo -> £2 
or Eq -> £3 are allowed, but with much lower probability now than the fundamental. 
They do in fact occur. The vibrational frequency of the molecule is still v, the v 
that can be calculated from the equation on p. 7. The only difference between 
molecules in any of the vibrational levels is the amplitude of the mode, Ar. The 
transition E —*■ £ 2 is called the first overtone, Eo -> £3 the second overtone. Both 
are indicated in Fig. 8. Furthermore, the smaller the x-value, i.e. the closer the approxi- 
mation to the pure harmonic oscillator, the smaller the probability. Since yY is 
proportional to the probability of a transition (/ being the intensity of an absorption 
band), the smaller the anharmonicity constant x, the smaller will be the absorption 
at the overtone frequencies compared with the fundamental. Conversely, intense 
overtones point to a large degree of anharmonicity of a vibration. 

For positive values of x, A£oi > A£i2 and therefore #02 < 2<7oi, i.e. the wave- 
number of the first overtone is less than twice that of its fundamental, whereas for 
negative x values the opposite is true. 

ROTATION 

Rigid diatomic molecule 

We start again with the diatomic molecule, masses mi and mz joined by a rigid bar 
this time, length ro. Such a molecule may rotate about the 3 axes through the centre of 
mass Z. Z is defined by mm = m^r^', see Fig. 9. Rotation about these axes produces 




--I 







Fig. 9. ro = n + r* 



12 THEORY 

three moments of inertia, I a , h and I c . The a axis coincides with the rigid bar (bond 
axis) and so the moment of inertia I a = 0. Furthermore h = h as the b and c axes 
are equal, h is denned as 

h = mm 2 + W2/"2 2 

or, with nun = m^r% and (x, the previously mentioned reduced mass 

lb = f*ro 2 

Now suppose the molecule has a permanent dipole moment directed to mi, as 
indicated in Fig. 10. Viewed from the left to the right, the molecule rotating in the 




Fig. 10. Movement of a dipole during rotation 

plane of the drawing produces an alternating dipole moment. Therefore such a 
molecule can absorb electromagnetic radiation provided its rotational frequency 
and that of the incident radiation are identical. Molecules without a permanent dipole 
moment cannot absorb radiation as they do not produce an alternating field during 
rotation. 

Just as vibrational energy is quantised, so also is rotational energy. From quantum 
mechanics it can be shown that the rotational energy levels for the diatomic rigid 
molecule are given by : 

E J I0t = r^7 J ( J + 1), where /=0, 1,2... 
J is called the rotational quantum number, while the selection rule is 

A/=±l 
The energy between two adjacent levels is thus given by 

2/z 2 
AEj^j+x^ = -—.(J+ I) 

OTT^lb 



or in wavenumbers 

2h 



oj-^J+V"* = 



(/+D 



%TT 2 I b C 

and since for a molecule h/Sir 2 IbC is constant the relation can be written as 

°j^j+i I0t = 2B(J + 1) 

where B is called the rotational constant. The levels are shown in Fig. 1 1 in cm -1 
(Ejhc), a common and useful method as any transition can be read off and searched 
for in a spectrum directly this way. A presentation as such is called a 'term scheme'. 
As A J = ±1 the absorption of radiation will take place at 2B, AB, 6B cm -1 etc., and 
hence the distance between two adjacent absorption peaks will always be 25 cm -1 , 
i.e. the peaks are equally spaced (see Fig. 14) whereas the energy levels are not. For 



13 

some molecules the value Ao' _ 1 is given in Table 1. As one can see these absorptions 
occur in the far infrared region. 

Absorption process. The absorption of electromagnetic radiation by a molecule 
and the conversion in rotational energy may be visualised as follows. 

Consider a non-rotating molecule, the centre of mass being fixed anywhere in 
space (no kinetic energy). Let the molecule be irradiated by monochromatic radiation, 
or in other words by a stream of photons. Suppose a photon reaches the molecule. 
Since the time of interaction between a photon and the molecule is very short - about 



i 


> 












50 








i 


i 


40 








(cm" 1 ) 








30 


- 








i 


i 


20 








10 




i 


L 




i 


i 




♦ 



1 



Fig. 11. Rotational energy levels for a rigid diatomic molecule. The arrows indicate the 
allowed absorption transitions 



10-is s- we w m h ave to trace the photon on its way through the molecule at 
extremely short time intervals. 

The photon is (creates) an alternating electric field and a molecule with a per- 
manent dipole will start rotating in order to follow the alterations of the field (cf. 
electric motor : a coil in an alternating field). As long as the molecule rotates in phase 
with the alternating electric field it will withdraw energy from the photon and return 
it when moving out of phase. 

The rotational energy of a molecule, given by E = %Ia> 2 , is as we know quantised, 
and thus also the angular velocity of the rotation, co, which is related directly to the 
frequency of the alternating electric field of the molecule itself. 

Two possibilities arise now. Either the frequency of the photon is identical to the 
first of the quantised angular frequencies of the molecule or it is not. The former will 
give rise to a complete absorption of the photon by the molecule because it moves in 
phase all the time the photon is 'present'. The latter may look complicated but is 
rather simple. Two waves, slightly different in frequency, will be more or less in phase 



14 THEORY 

during a certain period and out of phase for another (see Fig. 12). For the molecule 
(and the photon) this will lead to an absorption of energy in the first half-period, 
followed by an equal return in the second. The final result will be an undisturbed 
passing of the photon. In short, the molecule 'absorbs' the photon, tries to fit it 
into the first excited energy level and if it cannot, lets it pass again. 

For a molecule that is rotating already, a similar model can be set up. When a 
photon reaches such a molecule it will try to absorb the photon by increasing its 

first half -period 



completely completely 

in phase out of phase 

Fig. 12. Phase relationships of two waves 

angular frequency to the nearest quantised situation. If this frequency and that of the 
photon are the same, the photon will be absorbed. If not it will reject it for the same 
reason as mentioned before. 

Non-rigid rotator 

For a non-rigid bonding between mi and m%, as is usually the case, the distance 
m\-m% is no longer constant and hence the moment of inertia h is no longer constant 
either. The distance will be increased due to the centrifugal force, especially at high 
rotational velocities. The molecule may vibrate as well, but as the corresponding fre- 
quency is about a hundred times greater than its rotational frequency, one can make 
use of an average value for the varying distance r. 

Nevertheless the spacing between the energy levels is no longer given by the earlier- 
mentioned relation and has to be changed to 

<v-*./+i rot = 2B(J + 1) - 4D(J + l) 3 * 

where D is the centrifugal distortion constant, which is a positive quantity, roughly 
B x 10 -3 . The outcome can be seen in the term scheme (Fig. 13) and the corresponding 
absorption pattern (Fig. 14). As shown, the correction term proves to be without 
much effect for low / values. 

Intensity of rotational bands 

The intensity of a rotational band, belonging to the transition from level J to level 
/ + 1, is proportional to the number of molecules going over from /—»■./ + 1. 
The probability is the same for each transition J—*-J + 1 ; therefore the intensity is 
proportional to the number of molecules in a certain level, in our case in level /. 
Here one can make use again of Boltzmann's distribution law : 

Nj 



No = 6XP 



/ -A£(w \ 



AEcw = BJ(J + \)hc 
* Calculated fromF ro t = Erotlhc = BJ(J + 1) - DJ 2 (J + l) 2 



15 



J value 
10 



5 ■ 

4 

3 

2 

1 



Fig. 13. Term scheme for the rotational energy levels for a rigid (left) and a non-rigid (right) 
rotator 



1 


2 


3 


4 


5 


6 


7 


8 9 10 


▲ 


A 


▲ 


A 


A 


A 


A 


AAA 





1 


2 


3 


4 


5 


6 


7 8 9 



J transition 
non-rigid rotator 



J transition 
rigid rotator 



Fig. 14. Rotational absorption bands for a rigid and a non-rigid rotator. The wavenumber 
scale is arbitrary 



1 


2 


3 


4 


5 


6 


7 


8 


9 


10 


A 


A 


A 


A 


A 


A 


A 


A 


A 


A 





1 


2 


3 


4 


5 


6 


7 


8 


y 



where Nj is the number of molecules in level / and No the number of molecules in the 
ground state, J = 0. Obviously Nj decreases rapidly with increasing /, as shown in 
Fig. 15. 

Although we have always spoken about one level for each / it appears that each 
level has to be split into several completely identical levels; the number of these 
so-called degenerate levels is given by 2J + 1 . Hence the number of molecules 
in a certain level is proportional to the product of the ratio Nj/No and the 
number of levels 2/ + 1. Thus the intensity of an absorption band is proportional to 



16 



THEORY 




Fig. 15. The ratio Nj/Nq for various values of B as a function of the rotational quantum 
number /, at room temperature 



Nj/No (2/ +1) and, when this is plotted against /, curves are obtained as shown in 
Fig. 16 for certain B values. The peak with maximum intensity will be found at 



_ / 300 
" \2SSB 



Intensity of vibrational bands 

The intensity of a vibrational band belonging to the transition from n = -> n = 1 
proves to be proportional (by quantum mechanics) to the square of the change of the 
dipole moment for the corresponding normal vibration near the equilibrium position. 

In other words, for a linear diatomic molecule / 



molecules / is proportional to 



(dMy 



For more complex 



(8M x y [dM y y /dM,y 



with M x , My, M z being the three components of the dipole moment M of the molecule 
in the x, y and z direction in the displaced position of the atoms. Unfortunately the 
derivative of M is not known beforehand and so one does not know anything about 
the intensity of a band. From practice, however, it appears that in many cases the 
partial dipole moment, M par t, can be used as a rough approximation to the derivative. 
For instance, a C=0 group has a large M value and it absorbs strongly in the infrared, 



17 

while the C=C group in CH 3 CH 2 C=CCH 3 shows hardly any absorption, in accor- 
dance with the expected small M. For more complex vibrations M is difficult to 
determine. Experience can then be of great help. 

The intensity of overtones, n — -* n = 2, which are only possible for anhar- 
monic vibrations, depends on the anharmonicity term. The larger this factor the 
more intense the absorption band, though it is still much weaker than the fundamental 



10- 



5- 



B=l 



B-5 



BOO 



5 10 

J 

Fig. 16. Intensity of rotational absorption bands for various B values as a function of /. 
/ in arbitrary units 

absorption under identical sampling conditions. The CO group is a good example of a 
group producing an overtone (~3400 cm -1 ). Again one can make use of a rule of 
thumb : the stronger the fundamental the stronger the first overtone. 



VIBRATING ROTATOR 

From the preceding sections we know the energy transitions for a vibrating molecule 
to be about a hundred times greater than the rotational energies. Since these energies 
are so different we may, as a first approximation, consider that the molecule can 
vibrate and rotate independently. This assumption is a 'Born-Oppenheimer' like 
approximation, mathematically expressed as 

Emo\ = i^vib + -Erot 



18 THEORY 

Substituting the expressions for £Vib and E xo x as found before one obtains 

£moi = hv e [(n + i) - {n + i) 2 x] + hc[BJ(J + 1) - DJ 2 (J + l) 2 ] 
or in wavenumbers 

Fmol = a e [(n +*)-(« + i) 2 x] + BJ(J + 1) - DJ%J + 1)2 

where F = E/hc and the subscript e denotes an anharmonicity-corrected frequency 
or wavenumber. 

The term scheme for this expression is shown in Fig. 17 for n = and for n = 1. 
It may be shown that the selection rules for the combined motions are the same as 



- 


i 


i 


i 


' 







t 


L 


i 


k 






t 




i 


i 








t 


L ! 


i 


k 










i 


i 




ki , ^ 










F 


-ik— 








I — 1 i 

1 


r-f- 

i 

1 












°o 


























1 








































L_ 














i 










'r-^ 









4 n-1 



n=0 



(Q) P branch 



R branch 



Fig. 17. Term scheme for a diatomic anharmonic oscillator. Not to scale. In the P branch, 
m = J + 1, m < 0, A/ = — 1 ; in the R branch, m = / + 1, m > 0, A7=l 

those for each separately : AJ = ±1 and An = ± 1 , 2 . . . (for a diatomic molecule 
A/ = is not allowed except under special circumstances). 

The distance between two rotational levels, one belonging to n = and one 
to n = 1 is given by 

aGOo-i. = a + 2Bm — 4Dm 3 

Where m = J + 1 and m > for A7 = +1 
m < for A/ = - 1 

/w = ±1, ±2, ±3 . . . 



19 

Thus absorption bands may be observed at the frequencies indicated in Fig. 18. 
Bands at the low frequency side (A/ = —1) are referred to as the P branch, while 
those to the high frequency side (A/ = +1) are called the R branch. The band for 
AJ = 0, uncommon for diatomic molecules, is called the Q branch. 

According to the derived formula the spacing between two adjacent bands will 
be about IB cm" 1 for small m- values whereas it will decrease for increasing values of 
m (see Fig. 18a). 



R 
Branch 



P 
Branch 



Fig. 18. Distribution of the rotational bands superimposed on a vibrational transition for a 
diatomic molecule 




Fig. 18a. Rotational bands superimposed on the vibrational transition of CO at 2140 cm -1 

In fact the formula for the energy levels of a rotating vibrating diatomic molecule 
is somewhat more complex than the one which has been derived but we shall not go 
into more detail here. 



Molecular interaction 

Strictly speaking the formulae derived in the preceding sections are valid only for an 
isolated molecule. Actually the absorption of vibrational or rotational energy by a 
single particle cannot be observed. A much greater number of molecules is required, 
but some kind of interaction may then arise. 

Gas phase. Interaction between molecules will be lowest in this phase especially 
at diminished pressure. At room temperature the molecules are moving in all direc- 
tions and depending on the free path length they will collide regularly, e.g. for carbon 
monoxide about 5 x 10 9 times/second at 750 mm Hg and 300 K. At the moment of 
collision energy can be exchanged : one molecule may gain energy at the cost of another 



20 



THEORY 



one. The exchanged energy is either kinetic, or rotational, or vibrational or of com- 
bined character; see Fig. 19. 

The conversion of vibrational and rotational energy into kinetic energy is one way 
by which molecules dissipate their absorbed radiation. It may be observed as a rise in 
temperature. 

The conversion of kinetic energy (300 K) into vibrational energy is of no importance 
at wavenumbers above 400 cm -1 as the energy is too small to reach the first vibrational 
level. Since, however, the energy quanta required for rotational transitions are much 



Kinetic 



Kinetic 




Fig. 19. Possible energy exchange scheme 



smaller, these will be readily picked up from kinetic energy. At room temperature, 
therefore, many molecules are rotating, i.e. they occupy rotational energy levels 
(J ^ 0). Consider a rotating CO molecule, rotational frequency vj, at room tempera- 
ture. As a result of kinetic energy, collisions with other molecules will occur every 
2 x 10~ 10 s. Suppose a photon, frequency v J+ i, arrives just after a collision has 
taken place. Let v J+1 be slightly smaller than vj + 2B{J + 1). As mentioned before 
(p. 14), the molecule will 'absorb' the photon and see if it is a suitable one. It will 
take some time before it is discovered to be unsuitable, after which it will be rejected. 

If it has not been discarded before the next collision, the energy (i.e. the photon) 
may be transferred to the other molecule and hence it will look as though the first 
molecule has absorbed the photon in the normal way. 

The larger the interval between two collisions the better the control on the photon 
frequency, the smaller the deviation from the 'true' value and the narrower the absorp- 
tion band. 



21 

Let us take CO as an example 

IB = 3-84 cm- 1 and so 
v = 3 x 10 10 x 3-84 « 10 11 Hz 

Thus the frequency of a suitable photon has to be 10 11 Hz. The number of fully 
completed cycles between two collisions is 

1011 =20 



5 x 10 9 



and so the inaccuracy in the photon frequency is ^ or 5%. Hence collisions cause 
broadening of the 'absorption frequency,' or in other words, broadening of the rota- 
tional energy levels. As the frequency of a photon deviates from the exact frequency 
the probability of absorption decreases. 

Since the collision frequency decreases at reduced pressure, bands will then 
become sharper. For instance the bandwidth for CO at 75 mm Hg at room tempera- 
ture will be about ±0-5%. Conversely, the bands will become broader at higher 
pressure as collision frequencies will be increased. 

A similar consideration and calculation can be applied to the bandwidth of 
vibrational peaks. Suppose one has a peak at 3000 cm" 1 . The frequency will thus 
be 10 14 Hz. Provided the number of collisions per second is still about 5 x 10 9 , the 
number of completed vibrations in between two collisions is 

10 14 

= 2 x 10 4 



5 x 10 9 



So the inaccuracy in the band frequency will be ±5 x 10~ 3 % (±0-15 cm -1 , negligibly 
small in comparison with other causes of band broadening (see next section). 

Liquid phase. In the liquid phase, interaction between molecules is such that 
virtually no free rotation exists. Only the lowest Zirot levels play some role; the others 
will not be occupied. Besides, the lowest energy levels will be very broad due to the 
numerous collisions ; they even might overlap each other. The few bands are no longer 
separated but form one diffuse relatively broad band. In the infrared this will give rise 
to a broadening of the vibrational bands, for as we already know, the rotational levels 
are superimposed on the vibrational levels. Naturally these levels themselves are also 
spread, due to the increased number of collisions, but this is insignificant in compari- 
son with the broadening caused by the rotational levels. 

Summarising, we may say that on changing from the gas phase into the liquid 
phase, the rotational fine structure disappears and the vibrational bands become less 
sharp. A frequency shift for all bands is very likely ; it can be assumed to be due to the 
molecular interactions. 

Solid phase. In the solid state a molecule is fixed in a crystal lattice. Rotation of 
the molecule as a whole is therefore prohibited; rotational levels no longer play a 
role. The result will be a sharpening of the vibrational bands compared to the liquid 
state. (This does not hold for molecules crystallising in an ionic lattice.) 

Since organic molecules crystallise in general in unsymmetric lattices, a splitting 
up of bands may occur. Bands coinciding in the liquid state shift to different wave- 
numbers due to the fact that the corresponding modes are slightly different in the 
solid state structure. Even the same functional group (e.g. C=0) may show a split 



22 THEORY 

band as result of different orientation in the crystal lattice. It may therefore look as if 
two C=0 functions are present, but a solution experiment can clear up any 
ambiguity. 

Sometimes rather distinct bands appear which are not present in the solution or 
liquid phase. They may be attributed to vibrations of the lattice (partly or as a whole), 
as a result of, for example, the formation of associated molecules. 

Finally it must be pointed out that some bands might show a considerable orienta- 
tion effect (see p. 55) as result of instrumental polarisation, present in every spectro- 
meter. 

Triatomic molecules 

A triatomic molecule may be either linear or bent, the number of fundamental 
vibrations being 4 and 3 respectively. Apart from the stretching vibration, a bending 
mode is possible for these systems. This can be demonstrated best by some examples 
(see Fig. 20). The movements of the atoms in the plane of the drawing relative to the 



O-C-O - I 3312H-C-N 

£ "5 *■ •+ -* ► 

O-C-O 2349 H 2089 H - C - N 

A - C - A 667 m 712 AH - C - NA 

! ▼ 

+0 - C - 0+ 667 E? 712 +H - C - N+ 



A 

s O 3652 

H X H 

> *- \ 

/O x V 3756 

H H 

A A 

z O v 1595 

H H 

\ > 

Fig. 20. Modes of some triatomic molecules. The length of the arrows is not proportional 
to the real amplitude 



centre of mass are indicated by arrows. An arrow represents only half a movement of 
course; in the second period the direction is opposite. Perpendicular modes are as 
usual indicated by + and — representing movements towards and away from the 
reader respectively. 

Carbon dioxide. A linear symmetric molecule. Vibration I is called the totally 
symmetric stretching vibration and since there is no net alteration in dipole during the 
movement, this fundamental will be infrared inactive; no absorption band will 
appear. The second vibration (II) is called the asymmetric stretching vibration and 
since the dipole is altered, it will be infrared active. The next one (III) is a bending 
vibration, also active, while IV is the same vibration but in a plane perpendicular to 



23 

that of III. This will not give rise to another absorption band, however, since both 
vibrations are fully identical. This phenomenon is called degeneracy. 

Hence CO2 will only produce two fundamental bands; stretching and bending 
ones at 2349 cm -1 and 667 cm -1 respectively. 

Hydrogen cyanide. A linear non-symmetric molecule, the number of fundamentals 
still being 4. Number I (as well as II) is active now, due to the asymmetry of the 
molecule. Number I is called the CH stretching mode, while II is said to be the CN 
stretching mode. Here again as in the case of CO2, III and IV are degenerate modes at 
712 cm- 1 . 

Water. A symmetric bent molecule. Again I is called the symmetric stretching 
mode and II the asymmetric one. Vibration III is a symmetric bending mode. 
Movements out of plane will produce rotation only and can therefore be omitted. 
The spectrum of water will thus contain three fundamental absorptions (at 3756, 
3652 and 1595 cm- 1 ). 

More examples may be considered, but no matter what molecule is chosen, its 
vibrational modes belong to one of the three types represented here. 

Combination bands 

In addition to the fundamental and overtone bands, combination bands are found 
in the infrared region. The origin of such bands is rather simple. Consider a molecule 
with fundamental wavenumbers o\ and a 2 . It may also absorb energy quanta hcaz, 
where as is some linear combination of g\ and #2, i.e. 0-3 = 01 + #2, °"3 = 2cri + 02, 
a 3 = g\ — 02 etc. Which combinations are allowed and which are not can be deduced 
from symmetry considerations. Like overtones, these combination bands arise only as a 
result of anharmonic oscillation, for interaction between two vibrations is then possible. 

Degeneracy 

As we have seen before, the number of normal vibrations for a molecule is 3N—6 
or, for a linear molecule, 3N—5. This does not by any means denote the number of 
absorption bands that may be observed in the infrared. Whether a band does appear 
depends on several factors such as the selection rules, the intensity, the symmetry of 
the molecule and the infrared region under consideration. 

If two or three vibrational modes are fully identical as result of symmetry, they 
are said to be degenerate ; doubly degenerate as in the case of the symmetric bending 
vibration of CO2 and many other molecules, or triply degenerate. The latter will 
only occur in molecules with a high symmetry, and are of no importance in common 
infrared work. 

It may also happen that two or more vibrational modes of a molecule have the 
same(or nearly the same) difference in energy between two energy levels. The corre- 
sponding absorption bands, occurring at frequencies identical or very close to each 
other, will be indistinguishable. This phenomenon is called accidental degeneracy. 

Summarising, we may say that degeneracy causes a decrease in the number of 
absorption bands that may be observed. 

Isotope effect 

When a particular atom in a molecule is replaced by its isotope there will neither be an 
appreciable change in internuclear distance nor in the binding force. There is only a 



24 THEORY 

change in mass, and hence in the reduced mass p. Considering HCl as an example 
one can calculate the change in li to be about 0-15% in going from H 35 C1 (li = 0-9722) 
to H 37 C1 (it = 0-9737). Since B is inversely proportional to it, the rotational energy 
levels will be altered by 0-15% or about 003 cm -1 , obviously too small to be seen in 
the rotational spectrum. 

As the vibrational frequency for HCl is proportional to l/vV** the difference 
between the energy levels for both isotopic molecules will be about 007%. For an 
absorption at 2886 cm -1 this means a difference of about 2 cm -1 , sufficiently large 
to be seen. 

Since naturally-occurring HCl is a mixture of fairly large amounts of both chlorine 
isotopes, all absorption bands in which the vibrational energy of 2886 cm -1 is involved 
will appear as doublets, separation 2 cm -1 , provided spectrometer resolution is high 
enough. 

A more drastic change in the energy levels for HCl - the vibrational as well as the 
rotational - is obtained when the H atom is replaced by its isotope D ; the reduced mass 
increases from 0-97 to 1-89. As this is nearly a factor of 2, the rotational levels will 
be only half the distance apart (5 cmr 1 ), while the vibrational energy will be altered 
roughly by a factor of 1/^/2 or °' 7 - Hence the vibrational band for DO should be 
near 2000 cm- 1 . (It is in fact observed at 2100 cm- 1 .) 

For an isolated C— H molecule, ii is 0-92. The reduced mass for the 13 C— H group 
is 0-93, while it is 1-71 for C — D. From this it can be concluded that the change in it 
for a diatomic system will only be appreciable in the following cases: (a) a large 
difference in mass and (b) the lightest atom replaced by its isotope. 

The only other possibility arises when both atoms have about the same mass. 
For instance, H 2 (ji = 200), D 2 (li = 1-00), HD (li = 1-50), C— C (li = 6-00), 
C— 13 C O = 6-24) etc. It is apparent that the lighter the molecule the greater the 
isotope effect. 

Since the B value for most molecules is small (less than a few cm -1 ) the rotational 
energy levels will hardly be influenced by isotopic substitution, deuteration being 
an understandable exception of course in several cases. 

The vibrational energies are even less influenced, as the change is proportional 
to l/y>> but because of the rather high A£ V ib in comparison to AE TO t, the effect is 
far more pronounced. For instance, a change of 0-1% in li will give rise to a shift of 
1-5 cm- 1 for an absorption band at 3000 cm" 1 , visible under high resolution. 

For a C— H stretching band the substitution of H by D changes li roughly by 
a factor of 2. This will lead to a shift for this band from about 3000 cm" 1 towards 
2100 cm- 1 , (Tch/<tcd being proportional to V(/"cd//^ch) or roughly -y/2 en 1-4. 
In general the ratio ctxh/^xd will always be «* \/2 provided m x > w H . This effect can 
therefore be very useful for the identification of an X — H stretching band, on condi- 
tion that that particular H atom can be replaced by deuterium without any other 
changes to the molecule. 

POLYATOMIC MOLECULES 

An increase in the number of atoms per molecule causes a further increase in the 
number of normal vibrations. True, the number will decrease as a result of symmetry 
(infrared inactive vibrations) and (accidental) degeneracy, but the spectrum would be 
hard to disentangle if it had a random distribution of all these bands. Fortunately it 



25 

is found that compounds having a particular group in common do show absorption 
bands in the same region(s). For instance, all molecules with at least one CH3 group 
absorb at 2900 cm -1 . An explanation is given below. 

Functional groups 

Consider different molecules in which a CH3 group is present. If the vibrational 
transitions for a CH3 group as such are completely different from any other transition 
in the rest of the molecule, the transitions or the modes of the CH3 group will be 
independent ; the group seems to be isolated. Remember, however, it is the molecule 
as a whole that is involved in each vibration and therefore, though small, a slight effect 
is to be found. The result is that a CH3 group in a molecule absorbs in always the 
same narrow wavenumber regions (i.e. at about 2960 cm -1 ; see also the Tables). 

There are many more groups with rather unique vibrational energy transitions 
which behave like the CH3 group; for example the groups OH, C=0, C=N, NH2, 
C=C. They all give rise (if active) to certain absorption bands in well-localised regions 
in the spectrum. 

Such groups are called functional groups as it is usually possible to conclude from 
an i.r. spectrum whether the group is present in a molecule or not. 

Skeletal vibrations 

Many other groups cannot be used as such, their accompanying absorption bands 
being very much influenced by the rest of the molecule. A good example of this 
is the single C — C band. If for instance two C — C bands are coupled - a very common 
situation in organic molecules - there will be an appreciable exchange of vibrational 
energy between the two groups. 

This can be understood easily for, if free, both groups would absorb energy of 
the same frequency. 

The result of the coupling will be that the places of the C — C absorption bands in a 
spectrum are strongly dependent on the adjacent bands or in other words the skeleton 
of the molecule. Vibrations of this type are often referred to as skeletal vibrations, 
and are highly specific for each molecule. 

After this one may ask oneself what happens if two functional groups are attached 
to each other. Will they exchange energy strongly ? Let us take two examples. 

1. H3C — CH3. As the energy involved in the vibrations of a CH3 group is quite 
different from that of a C — C group, the latter acts as an isolator; there will be 
no appreciable exchange of energy between both CH3 groups. 

2. C=C — C=C. Here too the C— C bond acts as an isolator, but less efficiently 
than that in the case of the ethane molecule. This is due to the fact that it 
has a slightly double bond character. Some interaction between both C=C 
groups therefore takes place. Yet from practice it is known that in this case 
too, the C=C group can still be called a functional group, as its absorption 
bands appear in the expected region. 

Types of vibrations 

The spectrum of a molecule will consist of several bands. As far as these bands can be 
unambiguously attributed to certain well-known vibrational modes of functional 
groups, these have names. We will mention the most important ones for they are 

3 



26 



THEORY 



often used in the literature. Taking the methylene group as an example to start with, 
the movements and corresponding names are as indicated in Fig. 21. The notations 
underneath the names are common ones, though not the only ones possible. 

Similar normal vibrations can be visualised for other groups, e.g. NH 2 , NO2, 
CH3, etc. The same names are given to modes that are identical to those for the 
CH2 group. For instance v CBs s is the symmetric stretching vibration where all 
three H atoms move up and down along the direction of the C — H bond. v CH3 as is 





symmetric 


asymmetric 


symmetric 


stretching 


stretching 


bending 


s 


v as 


or 


V CH 2 


V CH 2 


scissors 
5 CH 



<y\/ \^ 



wagging 


twisting 




or 


W CH, 


torsion 



rocking 
P .CH 2 

T CH 2 

Fig. 21. Types of normal vibrations for a methylene group. O is the hydrogen, # is the 
carbon atom 



the asymmetric stretching vibration where two hydrogen atoms move upwards 
along the CH band while the other one moves downwards and vice versa. The others, 
<5 CH3 S , <5 C H 3 as and p CKs can be found in a similar way. 

The modes of an OH group attached to a carbon skeleton can also be indicated 
by the above-mentioned names; voh, the O — H stretching, v C -o the C — O stretching 
and (5oh, the /Ov bending vibration. 
C N H 

A sub-division of some modes such as 'out of plane' and 'in plane' is frequently 
used, whereby the plane of the molecular skeleton is considered to be the plane. 

LARGE MOLECULES 

Finally, some remarks will be given about very large molecules such as steroids or 
even larger ones like proteins and polymers. In these cases, the number of atoms 
per molecule varies from about 50 to 1000 or more. Will the spectra not be too 




T3 
G 



o 

o- 

X/l 










<u 



9* 



o 



OJ 



ft 

a 
O 



oo 



CO 

>, 

I— I 

O 



rnrniTn 



33NV1UWSNVU 



01 


♦6-869 


01 


Z9Z 


€ 


Zfr8 


e 


*Z-906 


8 


*O8Z01 


9 


1-6901 


fr 


*e-*9n 


fr 


*M8I1 


€ 


Z8Z1 


9 


ZZZ\ 


9 


C9ei 


01 


fr9frl 


01 


96M 


9 


*T-e891 


6 


*M09I 


e 


*9-I08I 


€ 


*OTZ8I 


e 


*0-frfr6I 


Z 


*Z-098Z 


OT 


frZ6Z 


6 


*T-ZZ0€ 


i 


UIO 



a 
o 



8 

3S 

e 

3 

I 



O 



2 



.. 6 

2 &H 



6 
6 

o 



5 5£ 



0) 



$ ® « '- 



T3 m-i 
<D O 

^ O 



<]) .a 
* c 



CD 



o 

o & a> ,3 a> 

Oj: x; is a 
(0 5h ow 



M-l C <I> 



CO 

Q) CO 

■— « u 

10 ^ u 



f0 



co co 



H g ;Q 43 g 



O t— « > +J H 111 H 



• o 



"»'M'N 'uopu<n ' luaotaiQ uouapjv '-pn <»S 5 u»P*»H 
pu»iBu3 ut psqsnqnd » pajUTJd 'S96t © 



ZT9 



CO 



o 



c 

x: 

PL. 



88SeSR9S^S 



SDNViJJWSNVUl 



01 


ozz 


6 


xez 


T 


888 


Z 


Z80I 


V 


zoei 


S 


ssei 


9 


ZLZ\ 


Z 


T8ei 


8 


99frl 


I 


S06I 


Z 


8Z0Z 


e 


9ZZZ 


9 


0Z93 


6 


ZQ8Z 


01 


668Z 


I 


0*99 


Z 


0269 


fr 


08Ifr 


fr 


06Zfr 


s 


0Z£fr 


I 


08ZS 


I 


UIO 











6 6 












c 
o 




£ £ 














m in 
















O .-1 


« 








<u 


6 


4> 
O 


n 

01 

« 

c 
J* 


.-4 o 

o o 








-.H 


C 

o 
o 




^— ** *-** 


E 






CO 


tlH 




«3 .Q 


« 







o 

c 
o 



S 

s 

8. 



i- 



o 

Oh 






■»'M'N 'uopuoi ' ju8d«»jo uou»PtV "PH uos V u»pA»n 
pu»i6u3 ui psqsnqnd 9 pajutjd 'S96I @ 



ZZS2I 



30 THEORY 

complex for analysis, as the number of fundamental vibrations is already about 1 50 
for the steroid? Besides, as steroids in general have no symmetry element, all these 
transitions are infrared active. Many vibrations, however, are to be ascribed to 
carbon-hydrogen modes and as such are observed in the same spectral region; 
they coincide or overlap and thus allow a simplification of the spectrum (Fig. 
22). For such a huge molecule as a protein the situation is worse. There are 
so many groups with a slightly different character and the skeletal modes are so many 
that hardly any distinct band can be observed in such a spectrum. The many bands, 
though of different intensity, do overlap in practically all interesting regions. The 
result is a spectrum of extremely broad bands in which, however, some fine structure 
(peak maxima, shoulders) can be found (Fig. 23). Obviously, these spectra cannot be 
used for structure determination, but they may be of great help in comparing or 
recognising proteins. 

Bearing in mind the spectra of proteins, one might be surprised to see that the 
spectrum of a polymer such as polystyrene (Fig. 24) is so simple, notwithstanding 
the large number of atoms per molecule. Here it is the simple composition of the 
polymer, in fact n times the same unit, that makes the spectrum look so simple 

— CH— CH 2 - 




The spectrum can be considered to be an «-fold superposition of the obviously rather 
simple spectrum of the molecule between brackets. The carbon skeletal vibrations 
will contribute also, but as their intensity is usually small, they will not seriously 
interfere with the simplicity of the spectrum. It is not surprising either that the spec- 
trum of polyethylene, (CH2)«, is even simpler than that of the molecule hexane, C6H14, 
as in the latter the CH3 groups are extra contributors (Fig. 25). 

BIBLIOGRAPHY 

G. Herzberg, 'Infrared and Raman Spectra of Polyatomic Molecules,' Molecular Spectra 
and Molecular Structure, Van Nostrand, New Jersey, 1945. 

K. W. F. Kohlrausch, Der Smekal-Raman-Effekt, Erganzungsband 1931-1937, Springer 
Verlag, Berlin, 1938. 



Spectrophotometers 



INTRODUCTION 

It is not the aim of this work to give in full detail the working and the construction 
of the normal type of i.r. spectrophotometers. The reader is referred to the literature 
as several books on this subject are available. The layout of a double beam spectro- 
photometer will be discussed using a block scheme : 



Source 


- 


Sample 
Compartment 




Mono- 
chromator 


- 


Detector 


- 


Amplifier 


Recorder 






























Wedge 
















sy 


sten 


1 



















The optical diagram of a spectrometer can be found over the page (Fig. 26). The 
different parts of this spectrophotometer will be dealt with in subsequent sections 
below, except for the sample compartment, which is described in Chapter 4. 

SOURCE 

The most ideal light source would be one emitting constant energy over the whole 
infrared region. Unfortunately as yet such sources have not been developed and so we 
have to do the best we can with the following ones : 

Nernst filament (ZrO and some other rare earth oxides) 
Globar (Si-C) 
Ni-Cr wire 
Heated ceramic 
Mercury lamp 

The main disadvantage of all these sources is the very unequal energy distri- 
bution in relation to the wavelength. The energy emission curve for 'black body 
radiation' is as given in Fig. 27. The curve is temperature dependent and can be 
calculated from Planck's formula 



h,T = 



2tt/2c 2 A- 5 



exp 



\kTXJ 



31 



32 



SPECTROPHOTOMETERS 



where / is the emitted energy, k is Boltzmann's constant and X is the wavelength in 
nm. The wavelength with maximum energy output is found from 

^max T = constant (Wien's law) 

whereas the total amount of emitted energy is proportional to T 4 (Stefan-Boltzmann 
law). 

The curves for the above-mentioned sources can only at best be as good as a perfect 
Planck curve. As one wants a constant energy output from the monochromator a 



PARABOLOID 
MIRROR 



ELLIPSOID MIRROR 




APERTURE STOP 



TOROID 
MIRROR 



Fig. 26. Optical layout of Perkin Elmer 157 



variable slit is necessary. The variation of the slit width has to be chosen in some way 
or another to be inversely proportional to the energy curve in order to be sure that a 
fairly constant amount of energy will reach the detector. To maintain a constant energy 
output the power supply for a source has to be stabilised. Fluctuations in voltage 
would cause fluctuations in the temperature of the source, and consequently changes 
in emitted energy. 

In most cases the Nernst filament as well as the Globar have to be connected in 
series with ballast lamps to limit the current. Both sources have a negative temperature 
coefficient; i.e. the resistance lowers with increasing temperature. To prevent them 
from melting the current has to be limited. 

MONOCHROMATOR 

The system of slits, mirrors and prisms and/or grating, necessary for the dispersion 
of the radiation into separate wavelengths, is called the monochromator. The infrared 



33 

beam passes from the entrance to the exit slit. The entrance slit can be viewed as the 
light source of the monochromator, its energy being fairly constant due to the varia- 
tion of the slit width. Plane mirrors are used mainly in connection with the splitting 
and recombining of the two beams. They can also be used to make an instrument 
more compact. Non-planar mirrors (spherical, ellipsoidal, parabolic) are used either 



(arbitrary 
units) 




2000 K(lxlO) 
1000 K(lxlOO) 



Fig. 27. Energy distribution for black body radiation (Planck's curve). Calculated for three 
temperatures from Planck's formula 



to produce a parallel beam necessary to obtain satisfactory dispersion with a prism 
or a grating, or to form a reduced image of the source in the sample area or of the 
exit slit at the thermocouple. This slit is variable also, thus determining the 'mono- 
chromaticity' of the passing radiation. In fact both slits are normally identical. 

Slit 

As seen before there are at least two slits, an entrance and an exit slit. The width of 
both slits varies simultaneously with the variation of the wavelength. This can be 
done either mechanically (by means of a cam) or electrically (by means of a non-linear 



34 



SPECTROPHOTOMETERS 



NaCI 5mm. 



5000 



KBr 5mm i 



CsBr 10mm i 



iCsl 10mm i 



LiF 5mm, 



CaF 2 5mm i 



AgCI 0-5mmi 



KRS5 2mm ■ 



Irtranl 2mmi 



i Ge 1mm i 



Irtran2 1mm 



Si O 2mm 



4000 



3000 



4%R 
8%R 
8%R 
9%R 
6%R 
3%R 
16%R 
19%R 
4%R 
38%R 
21 %R 
3%R 



2000 1000 200 

vfcrrT 1 ) 



Fig. 28. Regions of > 15%Tfor several materials. The thickness is as indicated. The data 
in the right column refer to % reflection losses 



35 

potentiometer). Most spectrophotometers have more than one 'slit programme', i.e. 
at any wavelength the slit width can be changed by choosing another slit programme. 
On more expensive spectrometers the slit programme can even be altered continuously. 
Obviously a smaller slit requires an increase in amplification. An increase in slit 
width causes a decrease of the resolution. This of course is also true for normal 
increase of the slit width at lower wavenumbers, but for prism instruments this is 
more or less met by the increasing dispersion of the prism in that region. 

Prism 

Prisms for i.r. instruments are made from rock salt (NaCl), KBr, CsBr, LiF, CaF2, 
KRS-5 etc. The prism material should be chosen in accordance with the resolution 
required. The dispersion of a prism depends on the change of the refractive index. 
The dispersion increases as the absorption band of the material is reached, though 
the n decreases rapidly. The transparency of the above-mentioned materials is shown 
in Fig. 28. 

Some of the prisms are sensitive to moisture. Protection is obtained by elevating 
the temperature of the prism some 20 degrees above room temperature. The tempera- 
ture has to be constant, as n, and thus the dispersion, changes with temperature. 

Grating 

Instead of the well-known prism as a medium of dispersion, a grating can be used 
for the same purpose. A grating is a mirror provided with numerous parallel lines or 
grooves. When a parallel light beam strikes such a grating, each line acts as a more 
or less infinitely small light source, emitting in all directions. Interference between 
the different rays from the different lines will occur. If the difference in path length 
between two successive rays of the same wavelength equals k times the wavelength 
(k = 1, 2, 3 . . . an integer) the intensity will be enhanced, whereas for (k + £) 
times, extinction will be the result (see Fig. 29). 
This mathematically expressed as 

d[sin p + sin (a + £)] = k . X 

where dh the distance between two parallel grooves, i.e. the grating constant, X is the 
wavelength of the rays looked at, and /5 and (/? + a) the angles between the incident 
and diffracted beam and the normal to the plane of the grating. 

For a given spectrometer the small angle a is constant since the entrance and 
exit slit are in a fixed position. Suppose now k = 1 (first order) then, since d is constant 
(inversely proportional to the number of lines per mm), the diffracted rays will form a 
spectrum because, for increasing values of (5, and thus sin /?, the wavelength of the rays 
will also increase. The spectrum will resemble one formed by a prism. 

A grating will, however, produce more spectra, e.g. a second one exists for k = 2 
(second-order spectrum) etc. Unfortunately these spectra do show overlap as can be 
easily understood from the formula: X in the first-order spectrum is overlapped by 
f A of the second order, by \X of the third order etc. A small prism or filter is therefore 
needed to separate one order from another. 

The intensity of the diffracted rays varies with both the angle /S and the order. 
The larger /? (and thus the wavelength) the smaller the intensity, and also the higher 



36 



SPECTROPHOTOMETERS 



the order the lesser the intensity. In general the first order is used as this gives the 
widest possible range of wavelengths together with less stringent filtering requirements. 

From the formula one obtains easily that for /? -► 90° the wavelength of the dif- 
fracted beam will be maximum, 2d = kX or X = 2d for k = 1 and a << /?. Such large 
/? values cannot be used for reason of lack of light yield. 

To improve the light yield in the first order at a distinct wavenumber, echelette 
type gratings are produced. The form of such a grating is drawn schematically in 




Fig. 29. The difference in path length between ray II and ray I is QB + BP or d sin 
(a -j- (!) respectively 



+ dsm 



Fig. 30. The angle <f> is called the blaze angle. It determines at what wavelength the 
light intensity will reach a maximum (/? & <f>). The grating is said to be blazed at that 
wavelength or wavenumber. 

DETECTOR 

The dispersed light passing the exit slit is focused with the help of a concave mirror 
onto a detector. In present-day instruments thermal detectors are used, i.e. the radia- 
tion is converted into thermal energy and the change in temperature is detected by 
either a thermocouple or a Golay cell. Several other detectors do exist but are still 
rarely in use at present (Bolometer, photo-tube etc.). 



37 




Fig. 30. Echelette type grating with blaze angle 4> 

Thermocouple 

Though specially constructed for its use in i.r. spectrometers, the action of the 
thermocouple is quite normal. A rise in temperature causes an increase in the electrical 
potential. Since the sample and the reference beam are focused onto the detector in 
turn the potential of the thermocouple will vary with time as indicated in Fig. 31, 
provided the sample is absorbing some energy. The period of this variation is equal 
to the time required for the chopper to complete one cycle. 



-Vo 




chopper 

frequency 

reference sample 

beam beam 

period period 



Time 
Fig. 31. Output signal thermocouple. (Reference and sample beam unequal) 



The output of the thermocouple is thus a d.c. potential Vo plus an a.c. potential 
of amplitude AK and frequency equal to the chopper frequency. 

The thermocouple is mounted in a small evacuated housing provided with an i.r. 
transparent window. The effect of the vacuum is to eliminate the loss of heat from the 
target to the air, resulting in about a four-fold increase in the temperature rise of the 
target in response to a fixed amount of radiation. 



38 SPECTROPHOTOMETERS 

Golay cell 

A Golay detector is a pneumatic detector; i.e. radiation is converted into heat, 
causing the expansion of a gas. The expanding gas alters the position of a mirror 
that forms part of an optical pathway ending with a photo-cell. The higher the 
temperature, the greater the expansion, the greater the distortion of the optical 
Golay system and the smaller the amount of light falling onto the photo-cell. 

The n.e.p. (see below) of this detector equals that of the thermocouple or may even 
be somewhat better. The Golay cell is more fragile than a thermocouple, however, 
and thus its lifetime is liable to be shorter. 

Noise 

The 'noise' of a spectrophotometer can be defined as the average random movement 
of the recorder pen, peak to peak, when both beams are unobstructed. It can be 
measured in % transmittance. It is primarily due to the 'Johnson noise' of the detec- 
tor - the small, spontaneous variations of potential which are an inherent result 
of its electrical resistance. By comparison, the noise produced by the amplifier and 
mechanical system can, in general, be neglected. (The equivalent to the Johnson 
noise, in the case of the Golay cell, is the thermal motion of the expansion gas.) 

A good measure of the performance of a detector is its 'noise equivalent power' 
(n.e.p.). This is the power (in watts) of the radiation incident on the detector necessary 
to produce a signal equal to the noise under certain standard conditions. A typical 
thermocouple has a n.e.p. of about 10~ 10 watt and a sensitivity of about 10 volts per 
watt. (Note: the lower the n.e.p. the better the performance.) 

Chopper 

Though the word chopper is adopted from single beam instruments (the radiation 
beam is interrupted (chopped) for half a period by a rotating disk), it is also used 
in double beam instruments for the mirror system that causes the reference and 
sample beam to fall alternately on to the detector. 

The chopper frequency has an upper limit determined by the response time of the 
detector. For a thermal detector, such as a thermocouple or Golay cell, this limit is 
typically between 10 and 20 Hz. In addition, to minimise interference from a.c. 
mains, the chopper frequency should not be near \\n times (n is an integer) the mains 
frequency or else should be synchronously \n times this frequency. 

AMPLIFIER AND WEDGE SYSTEM 

Most infrared spectrophotometers operate on the 'optical null' principle; that is, 
any attenuation of the sample beam by the sample is balanced by a corresponding 
attenuation of the reference beam by the 'wedge' (variable optical attenuator) so 
that both beams remain equal in intensity. When the two beams become unequal, 
due to a change in sample absorption, the detector produces an a.c. signal propor- 
tional to the difference. This signal is amplified and fed to a special electric motor - a 
servo-motor - which turns clockwise or anti-clockwise depending on whether the 
sample beam is more or less intense than the reference beam. The rotation of the 
servo-motor causes the wedge to move into or out of the reference beam until both 
beams are equal again. The a.c. detector signal then becomes zero and the servo- 
motor stops. Even with complete unbalance of the beams the a.c. signal from a 



39 

thermocouple is very small (about 0-1 /xV). This cannot be sent very far before ampli- 
fication is necessary, and this is performed in the pre-amplifier, after which it is sent 
to the main amplifier, which is tuned to the chopper frequency. 

The form of the wedge is such that its position in the reference beam is linearly 
proportional to the transmitted energy. Thus the recording of the position of the 
wedge as a function of frequency (wavenumber) amounts to the same as recording 
a spectrum. 

The coupling of recorder pen and wedge system may be either mechanical (built-in 
recorders only) or electrical, by means of a potentiometer coupled mechanically to 
the wedge. 

SCANNING CONDITIONS 

Introduction 

On each spectrometer, no matter how simple it is, one will find several adjustable 
and/or preset controls to operate the instrument. A manual will be helpful in setting 
them in the correct position for routine work, but one may have one's own wishes 
and ideas, and must therefore grow accustomed to the use of these controls in order 
to get the desired result. 

The infrared spectrum of a compound may present a great deal of information 
about the substance provided the instrument is used carefully. The spectrum is 
unique and specific ; each absorption band is unequivocally characterised by its intensity 
and its wavenumber, the former related to the number of molecules in the beam, the 
latter somewhat dependent on the sampling technique. To get the best spectrum we 
can, the spectrometer has to be adjusted to optimum condition. Many factors influenc- 
ing the overall result cannot be altered by the operator (e.g. optical path). Some 
factors have to be checked fairly regularly (e.g. wavenumber calibration, balance, 
zero and 100% line etc.), while others are to be chosen before the spectrum is scanned 
(e.g. scanning speed, slit programme, gain etc.). Several of these factors or instrumental 
parameters will be examined briefly in the following sections. 

Scanning speed 

The scanning speed is limited by the response time of the comb/servo system, a 
system that indicates the absorption by the sample. If the scanning speed is too high, 
the servo system will not be able to follow the absorption pattern caused by the sample. 
The absorption bands, especially those with a steep slope and of great intensity, 
will be deformed. Neither the intensity nor the wavenumber of these bands will be 
correctly recorded (Fig. 32). 

The optimum scanning speed is the maximum speed at which any decrease does 
not change the shape and the exact place of the absorption bands. Since 'time is 
money', it should not be chosen smaller than necessary. 

Some spectrometers are provided with automatic speed suppression (a.s.s.). The 
scanning speed can then be set rather high, for as soon as an absorption band appears, 
the a.s.s. causes a decrease in the speed during the recording of that peak. 

Correct settings of the speed and the a.s.s. can be found by trial and error. A 
speed reduction or a higher a.s.s. must not alter with respect to the shape and the 
position of the bands. 





O 



Oh 

at 

J3 

V) 

T3 
C 
«J 
JO 

c 
o 



■ft °- 

o " 

U. W) 

c 
'5 

G 
«$ 
o 



o 

c 

3 
C 



J8 








41 

Slit programme and resolution 

As we saw before, the slit width can be chosen between certain values. The narrower 
the slit the better the monochromaticity of the radiation passing the exit slit. The 
dispersion of the radiation itself depends on the construction of the monochromator 
(mirrors, focal length grating and/or prism etc.). For a monochromator the resolution 
or resolving power (R) is defined as vjdv, where dv is the difference in wavenumber 
between two wavenumbers that can just be distinguished by the spectrometer. The 
ultimate or theoretical R is equal to b dnjdX for a prism and to k . m for a grating, 
where b is the base length of the prism, dnjdX the dispersion, k the order of the grating 
and m the total number of lines. Obviously for a prism the limiting R varies with wave- 
length but it does not for a given grating in a given order. In practice R varies more or 





narrow _^ wide 



Fig. 33. Influence of the slit width on resolution and band contour. The gain was set optimal 
for (b) and kept constant 

less just as the slit width does. At a certain wavenumber the resolution can be changed 
by changing the slit programme from narrow to wide or vice versa. We have to choose 
the slit programme such that the instrument will produce the correct absorption 
spectrum. If the slit is too broad, resolution will be too low and the radiation will not 
be sufficiently monochromatic; bands or peaks lying close together will be deformed, 
and an average pattern will be the result (see Fig. 33). As we can see, bands become 
broader, intensities smaller, and details disappear. If the slit width is too small, 
however, bands may not show their true shape either unless the gain is appropriately 
increased (provided the limit of the gain is not reached and/or the resulting noise is 
still acceptable). 

In conclusion one may say that the choice of the slit programme is important. 



42 SPECTROPHOTOMETERS 

Gain 

The amplification of the signal, necessary to get the comb/servo system, and thus 
the recorder pen in motion, is determined by the setting of the gain control. Some 
resistance has to be overcome to get the system moving, i.e. without a minimum 
amplification the pen will be 'dead'. This minimum is related to the amount of radia- 
tion (and thus to the slit width) that reaches the detector. On the other hand there is a 
maximum amplification value also depending on the slit width. The higher the gain 
the faster the servo system moves to a correct off-balance signal. For too high a gain 
setting the servo system will pass the correct position, thus producing an opposite 
signal. It will therefore stop and return but it will pass the correct position and so on. 
This continuous movement of the servo or pen is called 'hunting', and the passing of 





Fig. 34. Influence of noise on the peak maximum 

the correct position is called overshoot. If the overshoot is not too severe the pen will 
eventually stop by itself at the correct position (damped oscillation). 

Another parameter connected with an increase in gain is the noise caused by the 
detector and electronics. The higher the gain the greater the noise and vice versa. 
For accurate wavenumber measurements a high noise level is undesirable as an 
absorption band distorted with noise is hard to record exactly (see Fig. 34). 

Luft formula 

In 1947 Luft* found the following empirical formula 

R*(SlN) 2 t = constant 
where R is the resolution of the instrument, strongly related to the earlier-mentioned 
slit width or slit programme, S/N is the signal-to-noise ratio related to the gain in 
connection with the noise (N), and t is the scanning speed. 

The formula is correct for most values of R barring very high resolution where 
the fifth power has to be used. 

Let us illustrate the formula by some examples. Say an increase in the resolution of 

* K. F. Luft, Angew. Chem. B19, 2 (1947). 



43 

a factor of 2 is required. This can be achieved at the cost of either a decrease in the 
speed by a factor of 16 or an increase in the noise - a worsening of the signal-to-noise 
ratio - by a factor of 4. Conversely a two-fold decrease of the resolution enables us to 
reduce the scanning time to yg °f tne original one. 

This reasoning is only applicable if the scanning conditions are chosen or set 
optimally. For example, a two-fold reduction of the resolution may be followed by a 
16-fold reduction in the speed, but it is not necessarily carried out; no coupling 
between the parameters exists beyond human intervention. Yet a smaller reduction 
is useless, while a greater one is inadmissible. 

The question arises as to how one should adjust the three parameters correctly. 
If one combination is known the others can be deduced from Luft's formula. To 
find one of the possible combinations, one may proceed as follows : 

1. Set the slit programme to normal; i.e. set the slit programme control to position 
N or to the recommended position; if there is an energy control instead set 
this one to normal or recommended energy. 

2. Choose the scanning time. The entire spectrum should be run in about 30 
minutes. 

3. Adjust the gain control for a suitable signal amplification. This can be done in 
two ways. Both tests should be carried out in a region where there is no atmo- 
spheric absorption, e.g. at 2000 cm -1 . 

Overshoot method. Adjust the 100% control such that the pen servo system 
indicates 90% transmittance. Bring an object into the sample beam until the trans- 
mittance is decreased to 80%. Withdraw the object at once and adjust the gain so that 
the pen shows an overshoot of 1%. 

Dead spot test. Place an object (wire grid) in the sample beam and/or adjust the 
100% control so the pen indicates about 50% transmittance. Block one beam and 
withdraw slowly. Repeat with the other beam. The pen should indicate the same trans- 
mittance in both cases (within the limits of the noise of course). If not, the gain should 
be re-adjusted. After this proceed as follows : 

1. Adjust the balance control. Block both beams and see that no drift or a very 
small upscale drift is present. Otherwise readjust. 

2. Run a test spectrum, for instance polystyrene. In general a sample showing 
broad and sharp bands as well as shoulders should be used. It must also possess 
some well-defined peaks (wavenumber accurately known). 

3. Next re-scan the spectrum at lower speed, e.g. half the original one. Compare 
the absorption ratios of a pair of bands in both spectra; preferably a sharp and 
a broad band lying close to each other. If these are identical the original setting 
was acceptable and perhaps even a faster speed is possible. This can be checked 
in the same way. If both spectra are not identical the scanning speed has to be 
decreased further. This procedure should be repeated until both spectra are 
essentially identical. 

If the instrument is supplied with other controls related to the scanning con- 
ditions, such as automatic speed suppression, a similar procedure is to be followed 
(see Potts and Lee Smith*). 

* W. J. Potts and A. Lee Smith, Applied Optics 6, 257 (1967). 



44 SPECTROPHOTOMETERS 

The settings once found can be changed later in accordance with Luffs formula to 
other combinations for special purposes. A final check as mentioned above is required, 
however. 

High resolution 

Spectra can be obtained by narrowing the slit to its smallest value at a certain wave- 
number. Say the slit width is decreased by a factor p, then the energy is decreased by 
/? 2 (two slits!) and so the gain has to be increased by/? 2 to compensate for the energy 
losses. The noise will be increased as well by the same factor /? 2 . In the case of instru- 
ments equipped with a time-constant control, the noise can be reduced using this 
control, provided the scanning speed is adequately reduced. The final choice will 
depend on the capabilities of the spectrometer and on the personal preference of 
the operator (e.g./? in the gain and thus still/? 2 in the speed). The final settings are to be 
controlled by running a test spectrum. 

Limited energy spectra 

If for any reason the background line shows a transmittance of about 90%, the ideal 
situation (background = 100% line) can be restored by placing into the reference 
beam a 10% absorbing accessory. It is found from practice that the background may 
have any value between zero and 100% transmittance and a variable absorbing 
accessory (reference beam attenuator) is useful. The restoring of the normal situation 
for double beam spectrometers by using an attenuator cannot, however, be done 
without repercussions. Suppose that a micropellet removes 60% of the available light. 
Now whether an attenuator is used or not, only 40% of the light is left and useful 
for absorption and detection purposes. Or, in other words, the available energy is 
decreased by a factor of 0-4. In order to get back to the original spectrometer settings, 
the gain or the slit or the speed are to be changed. If the light losses are very serious 
(ultra micro work, beam condensers), e.g. up to 90% or even higher, all three con- 
ditions have to be altered: 

1 . The speed as slow as possible. 

2. The slit programme as wide as possible. 

3. The gain as high as tolerable/useful with respect to the noise. 

Even then the spectrometer settings may not be optimum and a slightly distorted 
spectrum may be the result. The usefulness of the spectra obtained in that way have 
to be controlled by comparing with spectra from the same compound run under normal 
(routine) conditions, but a somewhat distorted spectrum is often better than none ! 

Stray light 

Light of a wavenumber j£ a reaching the detector simultaneously with light of wave- 
number a is called stray light. Provided the stray light is chopped (has passed the 
chopper) it will be amplified in the same way as the normal light beam signal. A 
false or partly untrue signal will be the result. Stray light is only permissible as long as 
its energy is very small; less than 0-5% in comparison to the energy of the real radia- 
tion. Stray light is the result of unwanted reflections and marginal rays just missing 
the optics. To minimise the stray light a monochromator is divided in compartments 
separated by black painted screens and if necessary provided with small holes. 



45 

Nevertheless some stray light does exist. The wider the sUts the more stray light is to be 
expected. As we saw before the energy distribution over the complete wavenumber 
region is anything but flat. An energy ratio of 200: 1 for 4000 and 600 cmr 1 is quite 
normal. Hence it is clear that the stray light problem will be most important in the low 
wavenumber region: the slits are widest there and about 0-1% stray light of 4000 cm -1 
causes a 20% error. 

For qualitative analysis a total stray light percentage up to 2% is admissible. 
A straightforward reduction is possible by using special filters and/or double mono- 
chromators. More accurate quantitative measurements can be then carried out. 



Automatic gain and slit control 

In double beam instruments the energy falling onto the detector is not the same 
for all wavenumbers, despite the moving slits. For instance, in the regions where 
atmospheric absorptions are present (CO2, H2O) the remaining energy, if any, is low. 
The same holds for the regions where solvents absorb when the compensating tech- 
nique is followed. 

As the signal from the detector is linearly amplified, the final signal, the one 
supplied to the servo system, will vary greatly with the wavenumber. Having set the 
gain control to an optimum value at a certain wavelength (100% transmittance !) it 
will be wrong at all wavenumbers where atmospheric or solvent absorptions take 
place. The pen servo system will be slow to respond there, and a false spectrum 
may be the result. This problem can be overcome by changing either the slit width 
or the gain in those regions. Some spectrophotometers are already equipped with 
such devices, called automatic gain control (a.g.c.) or automatic slit control (a.s.c.) 
respectively. 

In the case of a.s.c. the energy of the reference beam is kept constant by altering 
the slit width when passing an atmospheric or solvent band. Unfortunately resolution 
will be altered simultaneously. In the case of a.g.c. the gain is usually adjusted for 
minimum energy (at the wavenumber where atmospheric or solvent absorption is 
highest). For higher energies during scanning the a.g.c. system will then reduce the 
gain. Of course such a system produces variable noise in the spectrum ; low noise at 
high energies and vice versa. Nevertheless it may be of great help when 'difficult' 
solvents are to be used. 



Scale expansion 

Expansion of the abscissa is frequently referred to as scale expansion. Though 
strictly speaking this may be permissible, the expression 'scale expansion' is to be 
restricted to an expansion of the ordinate. Few instruments are equipped with such a 
built-in unit, though some can be supplied with such an accessory. 

This feature can be used when only a very small amount of sample in too low 
a concentration is available. The resulting spectrum will normally be too 'thin', 
i.e. the absorption pattern will be recorded between 100 and 90% transmittance. A 
5 times scale expansion will now extend this region to 50 %, while for 10 times expan- 
sion % transmittance will be reached again. 



46 SPECTROPHOTOMETERS 

At first glance this may look promising and simple to do, the expansion being car- 
ried out either electrically or mechanically. At least two disadvantages are readily 
found, however: 

(a) The noise is expanded as well by the same factor. For instance, from 0-2% to 
2% for 10 times expansion. 

(b) When the background does not coincide with the 100% line (a normal situa- 
tion), the difference between both lines is expanded as well. 

This is disastrous for 'background lines' showing a slope from 50% at 5000 cm" 1 to 
100% at 600 cm -1 , or more. The maximum expansion factor is now 2, as can be 
easily verified. Since the situation is usually worse for solids (due to reflection losses 
etc. especially at high wavenumbers) the expansion factor is limited for that reason. 
Yet scale expansion by moderate factors can be useful, for time is saved if the concen- 
tration or the cell thickness is too low. 

Calibration 

Calibration of an instrument should not only be done when installing the apparatus 
but also fairly regularly thereafter. There are two important things to be checked 
very carefully: the accuracy of the transmittance scale and the accuracy of the wave- 
number (frequency or wavelength) scale. Other tests can be carried out, but these may 
be different for all instruments. 

Transmittance scale. This scale can be easily checked by using in the sample beam 
fast-rotating sectors of known transmittances. A high speed of rotation is necessary 
(up to 2000 r.p.m.) to avoid interference between the rotating sector and the chopper 
motor frequency. 

If the indicated transmittance is false, this can be restored by a re-adjustment of the 
comb, source or thermocouple though this latter can be a difficult and laborious job. 
For errors beyond the tolerated limit the instalment of a new comb is required. A 
periodical check is desirable, say about once a month when intensity measurements 
form an important part of the work and provided there is no significant change during 
this period. 

For qualitative measurements, such as structure elucidation, the periodical check 
can be less frequent, say once a year. 

Wavenumber /wavelength scale. The wavenumber or wavelength check has to 
be done at least once a week and even once or twice a day if very accurate measure- 
ments are to be performed. For this purpose the rotational bands of gases such as 
CO, CEU, H 2 and NH 3 are used. The gas is let into a gas cell (path length 10 cm) 
and the spectrum is recorded under certain scanning conditions. Very accurate data 
for these rotational bands may be found in the literature (see Appendix A). 

When the utmost accuracy is not essential, other more convenient materials may 
be used, such as polystyrene and indene. False wavenumber indication (i.e. error 
exceeding the tolerance) can be eliminated by re-adjustment. The prism-littrow system 
or the grating and the wavenumber scale indicator or marker should be disconnected. 
The exact position should be found by trial and error and then the systems should be 
reconnected. 

The deviation can be either systematic (all wavenumbers being somewhat too 
high or too low) or variable. In the former case the error can be overcome by the above 
mentioned re-adjustment procedure. 



47 

In the latter case some misalignment in the optical path of the instrument must 
be present. This can be corrected best by a service engineer. It is not always necessary 
to correct small deviations (a few wavenumbers or a hundredth of a micron for 
example), as one can compensate for them provided the correction factor is known. 
If the spectrometer uses pre-printed sheets of chart paper, it is even simpler to insert 
them so that the error (if a systematic one !) is corrected directly. 

Recording paper. The recording paper is a source of misunderstanding. One can 
distinguish two different types of recording: on loose pre-printed charts or on rolls of 
unprinted paper. 

In the former case the size of a spectrum is completely fixed. For most spectro- 
meters, however, these sizes are too small in view of the accuracy with which the wave- 
numbers can be determined. The exact size can be easily calculated as follows. 

Suppose the average accuracy to be 4 cm -1 . If a spectrum is produced from 4000 
to 600 cm -1 about 3400/4 clearly distinguishable positions are required to define the 
total amount of information. Furthermore, supposing that each position requires 
1 mm, which seems to be fairly reasonable, the length of the spectrum should be about 
850 mm. 

Since the accuracy of the transmittance is not better than 1%, the number of 
positions is 100 and the vertical dimension should be therefore at least 100 mm. It is 
convenient however, in view of the overall image of a spectrum, to use a larger vertical 
dimension. From the above calculations, though rather rough, it is obvious that many 
pre-printed charts are too small. 

Spectrometers recording on unprinted, blank paper in general can produce spectra 
of good dimensions. The only disadvantage, but a very important one in service work, 
is the impossibility of reading the wavenumber of a band straight off. Instead each 
band has to be determined by making use of some type of marking mechanism, which 
in its turn has to be calibrated on polystyrene or gases. With this type of spectro- 
meter expansion is easily done. 

With increasing automation flow chart recording on pre-printed paper is also 
possible now. This enables us to use either pre-printed charts with the advantage of an 
easy and exact read-off, provided the dimensions are right, or unprinted sheets with 
the possibility of expansion of the abscissa. A marking accessory is usually available on 
request. 

Paper shrink and stretch. Due to changes in humidity or temperature chart paper 
is subject to shrink and stretch. For 'ready-to-hand' spectra this is unimportant, as 
well as for unprinted sheets or notes. The situation is different for pre-printed charts, 
since stretching or shrinking of these charts before the recording of a spectrum may 
result in false registration. A check on this point in advance is advisable. 

BIBLIOGRAPHY 

A. E. Martin, Infrared Instrumentation and Techniques, Elsevier, Amsterdam, 1966. 

For calibration, see 'Tables of Wavenumbers for the Calibration of Infrared Spectrometers,' 

Pure Appl. Chem. 1, 4 (1967). 



4 
Sampling 



The appearance of a spectrum depends to a large extent on the preparation technique 
that is followed. Gas, liquid and solid spectra differ a great deal, as one knows from 
theory. Different spectra can, however, also be expected as a result of the preparation 
technique, the pre-handling of the compound and last, but not least, personal experi- 
ence. What is really required is a spectrum containing as much information as possible. 
No general rule or rules can be outlined to achieve this, but a survey of the most com- 
mon techniques will be of some help. 

The different techniques are discussed briefly in the following sections, though no 
procedures are given. More details are to be found in the literature and/or handbooks. 

GAS 

A gas is most simple to handle; no mixing with other materials is required and thus it 
is easy to remove. Of the three phases it is however the one with the lowest density 
and consequently a rather large volume is required to obtain a reasonable spectrum. 

A gas cell usually consists of an evacuable space supplied with two infrared trans- 
parent windows and at least one valve. Common dimensions are a path length of 10 
cm and a volume of about 170 ml. It is evacuated first and then a gas is let in. The 
amount of gas in a cell can be diminished by lowering the pressure. The main dis- 
advantage is the rather high 'dead' volume; only a small part of the gas is irradiated 
and thus useful for measurements. This can be overcome by constructing a cell 
that is tightly fitting round the radiation beam. Such cells are called minimum volume 
cells. They are commercially available with a path length of 7-5 cm and a volume of 
25 ml. They are as easy to handle as the normal gas cell. Yet the volume in relation 
to the path length is a serious drawback for this type of cell. 

A great step forward is the multi-pass cell. In these cells a considerable increase in 
path length is obtained with the help of a couple of inside mirrors. The radiation is 
reflected that way several times without the volume being altered. Path lengths of 1 m 
or even about 20 m are then possible. As result of the many reflections, however, 
the transmittance is decreased to 30% or less (due to stray light and reflection losses) 
and so the signal-to-noise ratio is influenced unfavourably. Other disadvantages are: 

(a) The double-beam character of the spectrophotometer is completely distorted 
when a multi-pass cell is used. Water and carbon dioxide bands will no longer 
be compensated for. 

(b) The cell is difficult to install, and minor changes in the adjustment cause a 
dramatic decrease in the transmittance. 

48 



49 

(c) Although the efficiency for a multi-pass cell with respect to the gas volume is 
much higher than for an ordinary cell, it has been found to be necessary to 
increase the volume for the construction of cells with large path lengths. 
Therefore the absolute amount of gas required to fill a cell is increased. 

A spectrum of a gaseous compound may also be obtained in solution. This 
method can be of importance if only small quantities of a gas are available in an excess 
of others, such as air. By dissolving the gas in a suitable solvent it is also concentrated 
and N2 and O2 do not interfere. The spectra of course are no longer gas spectra, 
but have great similarity to liquid spectra. Obviously the absorption of the solvent is 
a serious drawback of this technique. Furthermore, this method is more laborious 
than the preceding ones. 

LIQUID 

The liquid compound as such can be transferred into a liquid cell of fixed path length, 
10 to 25 /mi, by means of a syringe or, even simpler, by adding a few drops to one 
of the supplying channels. The disadvantages of this method are numerous. 

(a) Such thin cells are hard to clean in many cases. 

(b) The cells have rather high dead volumes, viz. the supplying channels and the 
non-irradiated part. 

(c) The path length may be too large in some cases and too small in others. 

(d) The recovery of the substance is difficult. 

The amount of liquid can be minimised by making use of minimum volume cells, 
microcells or cavity cells. These have no longer an important dead volume, but they are 
still difficult to clean, especially when viscous liquids are involved. 

The easiest way to handle a liquid is to introduce a few drops between windows. 
If the path length is important a spacer can be used, but it may be left out equally well. 
In general, the thickness can be adjusted by altering the pressure by which both 
windows are pressed together. The window technique cannot be used with too 
volatile compounds nor with unstable ones. Contact with air, though limited to the 
outer area, can be ruinous. This method is simple, however, and cleaning can easily 
be done afterwards. For quantitative measurements liquid cells with larger path 
length are to be preferred. The compound has to be dissolved; the spectrum is that of 
a solution. For cells of 0-1 and 0-5 mm a concentration of 5 and 1% w/v will usually 
do. Cleaning of such cells is no longer a problem. A limiting factor of course is the 
solvent's own absorption, despite compensation techniques. Therefore at least two 
different solvents, e.g. CCU and CS2, are required to cover the normal i.r. range, 
2-16 /urn. 

Volatile or fairly volatile liquids may be transferred to a gas cell. This can be 
more useful than liquid cells as gas bubbles form easily and are difficult to remove. If 
the vapour pressure is too low to obtain reasonable spectra an increase in temperature 
might be helpful. 

SOLID 

A solid compound can be dissolved in a suitable solvent, after which the solution is 
transferred into a liquid cell. The disadvantages are as mentioned above for liquids. 



50 SAMPLING 

Moreover the choice of an acceptable solvent, taking into account the solubility of the 
compound and the absorption of the solvent, will be very difficult or even impossible. 

The mull technique is another common way to obtain the spectrum of a solid. 
Finely powdered material is homogeneously mixed with an inert, fairly transparent 
liquid such as paraffin oil. This is done to overcome the disastrous amount of scattering 
of light that would occur if the powder was brought into the sample beam between two 
NaCl or KBr windows. Disadvantages are obviously the absorption of the mulling 
agent, despite compensation techniques, and inhomogeneity due to gravitation. The 
technique is to be recommended, however, if one is interested in the absorption pattern 
of the crystal structure of the solid compound. 

The most extensive information about the crystal structure is obtained if one can 
prepare a very thin plate, 2 by 10 mm and about 10 ^m thick. Even polarised light 
can then be used. Unfortunately this technique is restricted to a few cases only. In 
general no such crystals can be produced. 

The film technique is applicable in several cases. The compound is dissolved in a 
volatile solvent. The solution is applied onto an NaCl or KBr window in drops after 
which the solvent is evaporated. A clear transparent film may be the result. Fatty 
compounds can be easily handled in this way. Most compounds however produce 
opaque films. For low-melting materials the preparation of a film via the liquid state 
can be tried between two windows. Films generally have a polymorphic or amorphic 
structure. 

KBr is transparent up to about 28 pm. KBr powder, mixed with a few mg of the 
unknown solid, can be made transparent by increasing the pressure to about 10000 
kg cm -2 and moulding a disk of the mixture under this pressure. 

The concentration in a KBr disk should be about 0-3%. Although this technique 
would seem to be ideal, there have been found to be serious drawbacks in practice. 
One has to be prepared for phenomena such as polymorphism, chemical (exchange) 
reactions, sensitivity to moisture, adsorption etc. 

Furthermore, the technique requires some experience before useful transparent 
disks are obtained. 

Of course numerous variations are possible with these techniques. One has to 
bear in mind that in fact each problem demands its own technique, and so one has to 
be familiar with the advantages and disadvantages of all possible ones. 

POLYMORPHISM 

The occurrence of several crystal forms of a compound is called polymorphism. Each 
form will show its own absorption pattern, differing slightly or significantly from the 
others. As this might give identification problems many workers prefer the solution 
technique. The molecules in a solvent will be 'unaware' of their origin and thus will 
give rise to one spectrum only. The disadvantages are clear: the problems concerning 
the solvent and the loss of information about the crystal structure. The KBr technique 
overcomes both drawbacks at once, but unfortunately may introduce other new 
problems. It is found that several compounds show alterations in the crystal 
lattice as result of the grinding procedure and/or the applied pressure. Changes fre- 
quently occur in organic materials having a low melting point (< 100°C), the com- 
pound being rendered completely amorphous. As one does not know beforehand 
the behaviour of the substance during preparation, one has to be very attentive to 



51 

polymorphic phenomena. As long as only one form is present each time - no matter 
what changes in the procedure - there is no need to discard the KBr technique. 
A run using a mull can lead to a decisive answer. Polymorphism has been observed 
for benzil, succinimide, several barbiturates, diacetamide, steroids etc. 

SOLVENT EFFECTS 

The spectra of the same compound in different solvents usually show differences. 
These are caused by the interaction of the solvent molecules (excess) and the molecules 
of the sample, and are therefore called solvent effects. Bands may be shifted to other 
wavenumbers and intensities can also be altered. The frequency shifts may be con- 
siderable; a shift of 100 cm -1 for a carbonyl group in going from one solvent to 
another is not uncommon. Many theories have been developed to describe the 
interaction and to predict the shifts, but so far none really fits the facts. The polarity 
of the solvent seems to play an important role. 

Comparison with reference spectra of solutions is difficult for this reason, without 
taking into account the solvent's own absorption regions. Structure elucidation is 
thus hampered by the solvent effect. The conclusive proof of an unknown and a 
reference spectrum being the same is only possible if all scanning conditions including 
the solvent are fully identical. 

Hydrogen bonding 

The attractive force existing between a slightly positively charged hydrogen atom on 
one hand and a rather negatively charged atom on the other hand is called hydrogen 
bonding. The hydroxyl group of methanol and the carbonyl group of acetone can be 
taken as an example : 

CH 3 — O x /CH 3 

x h o=c<; 

X CH 3 

The hydrogen bond or bridge is represented by the dotted line. The — OH group 
is the proton donating group, the C=0 group is called the proton acceptor. Another 
proton donator is the — NH2 group. Other proton accepting groups are the halogens, 
— N(CH 3 ) 2 etc. 

The attractive force of the bridge is small in comparison with covalent bonds; 
its absorption bands are to be found beyond our scope in the far infrared. Neverthe- 
less in the normal i.r. region the results of the bond are readily found. It appears 
that in the formation of the bridge both of the other bonds are involved. In our 
example the O — H stretching vibration as well as the C=0 are shifted to lower wave- 
numbers. 

Not only can a bridge be formed between different groups, but also between 
identical molecules. For instance, two alkanol molecules may form a dimer: 

,- H \ 
R— 0< x O— R 

X H'' 

Trimers, tetramers and polymers are possible as well. If one has simultaneously four 
different types of bridged molecules, then the spectrum will be composed of four 



52 SAMPLING 

different spectra, each belonging to one of the polymers that are present. A very 
complex spectrum would be the result, but fortunately the hydrogen bond only 
influences the vibration of the donor and acceptor group and hardly any of the rest 
of the molecule. 

Usually hydrogen bonding, especially O — H bridges, can be easily detected from 
the specific shape of the bridged group (see p. 76). Instead of the sharp band belong- 
ing to the unbridged (free) OH group one gets a rather broad band at much lower 
frequency. The broadness of the band is due to the fact that the bands belonging to 
the different polymers lie so close together that they cannot be resolved. Even under 
high resolving power the bands will not be separated, for they overlap inherently. The 
differences between the trimer, the tetramer and the other higher polymers are too 
small to be seen. The free OH and the dimer will, however, show well-separated bands. 

Intermolecular bridge. As long as the hydrogen bridge is the result of two molecules 
attracting each other the problems of interpretation of a polymer spectrum can be 
avoided by working in a dilute solution. For as one is concerned with equilibria such 
as 

n ROH ^ (ROH)n 

where n = 1, 2, 3 ... it is obvious that in a very dilute solution only the 'free' ROH 
molecule will be present, the spectrum being then rather simple. The lower the con- 
centration, however, the larger the path length of the cell, and thus the greater the 
distortion of the spectrum caused by the solvent. 

The concentration at which only 'free' molecules are present depends of course on 
the equilibrium constant; the larger the K the smaller the concentration. This effect 
may be present to such an extent that the rest of the spectrum is completely obscured 
by the solvent's own absorption. Yet some bands, e.g. a free OH band, can nearly 
always be located, since there are several solvents available, e.g. CCU, CS2, which 
do not absorb in the OH stretching region, even at path lengths up to 10 mm. 

Intramolecular bridge. In several molecules the donating and the accepting group 
are both present, for instance o-chlorophenol : 

/H 





(a) (b) 

Since the bridge is part of the molecule itself, it is called an intramolecular bridge. 
Contrary to what was said about the intermolecular bridge, diluting the solution will 
have no effect here. The bridge is permanent, there being no equilibrium similar to 
that for the intermolecular situation. However, an equilibrium does exist in this case, 
as part of the molecules will have the same configuration as molecule (b). The bridge 
has been broken in this case as result of kinetic energy. So in fact there must be an 
equilibrium constant depending on the temperature. The weaker the bridge and the 
higher the temperature, the higher the number of 'free' molecules. The equilibrium 
constant for the intermolecular bridge is also dependent on temperature ; the lower it is, 
the greater the number of molecules in the bridged form. 



53 

In the case of an intramolecular bridge dilution may have some effect. This is 
easily understood if one assumes that at higher concentration an intermolecular bridge 
is present as well: 

CI 

H \ 

— o o- 





\ v 

\ / 

Cl 

This is possible provided the O H bridge is weaker than the H Cl bridge, otherwise 
no intramolecular bridge would have been detected in dilute solutions. 

Whether a hydrogen band has inter- or intramolecular character can be traced 
by running spectra of dilute solutions. If the band belonging to the bridge is sensitive 
to dilution the bonding is intermolecular, and at low concentrations a sharp 'free' OH 
band will appear. If this does not happen, intramolecular hydrogen bonding may be 
present as well. Alterations in temperature may give further evidence. It must be 
emphasised that the choice of the solvent is very important in hydrogen bridge 
measurements. Interaction of the compound under survey and the solvent must be 
considered to be fully absent. The solvent has to be very pure, and any traces of water, 
a strong hydrogen bonding agent, must be totally removed. 

STRUCTURAL ISOMERISM 

Structural isomerism is frequently met in infrared spectroscopy when dealing with 
liquids, solutions or gases. Consider as an example the molecule 1,3-dichloropropan- 
2-one. Rotation about the C — C axes is possible provided energy barriers of a few 
joules/mole can be overcome. Kinetic energy at room temperature is ample for this 
purpose. Therefore at least three conformations of the molecule are possible : 



H Cl 




H 


Cl 


Cl 


H 




\ / 






/ 




/ 




C 




C 


< 


C 




h'\ 




4 

H 


\ 


H 


\ 




C= 


=o 




c=o 




c= 


=o 


\ / 




H 


/ 


H 


/ 




c 




< 


z 




C 




• \ 




• 


\ 


V 


\ 




H Cl 




Cl 


H 


Cl 


H 





(a) (b) (c) 

In (a) both chlorine atoms and the carbonyl group are in the C — C — C plane. In 
(b) just one chlorine atom is in that plane while that of (c) has no chlorine atoms in it. 
Since those conformations are optically different three types of spectra will arise. 
Although many normal vibrations will be identical for the three isomers (absorption 
will take place at the same frequency) some will be clearly different, giving rise to 



54 



SAMPLING 



bands at other frequencies. In this case, for instance, three carbonyl bands are found 
at 1755 (a), 1742 (b) and 1728 cm" 1 (c) respectively, indicating interaction between 
the chlorine atoms and the C=0 group. 

The relative distribution of the isomers will depend to a large extent on the 
energy barriers that have to be overcome. The intensity of the bands might give 
some idea of this distribution. A change in temperature will in general alter the ratio, 
while solvent changes might cause dramatic shifts, both in frequency as well as inten- 
sity, interaction with the solvent being the underlying factor. 

Sometimes the heights of the energy barriers are such that no equilibrium at room 
temperature exists ; the different isomers can then be isolated. For instance this is the 
case with many steroids. Infrared spectroscopy can be an effective tool in distinguish- 
ing between the different isomers. The spectra of the eight isomers of 5a,/?-pregnane- 
3a,/?-17a,/?-diol proved to be clearly different from each other. An unknown compound 
may be readily identified that way. Whilst it is true that structural isomers interfere 
with the infrared technique, it is thanks to that very interference that infrared spectro- 
scopy can give information about conformational problems. 

CHRISTIANSEN EFFECT 

The Christiansen effect is restricted to solid samples only. It is caused by a significant 
difference between the refractive index of the sample and the surrounding material, 
such as KBr, in the region of an absorption band. 




Absorption 



Fig. 35. The change in the scattered light and refractive index when passing an absorption 
band. 'True band', scattered light, «sampie, WKBr 



55 

Suppose «KBr A < Hsampie* (which is very common). In the region of an absorption 
band of the sample « K Br will be fairly constant, but the refractive index of the sample 
changes dramatically when passing the band (see Fig. 4). As the amount of radiation 
lost by scattering is proportional to the second power of the difference («KBr — ^sample), 
the background line will have the shape as indicated (Fig. 35). Compared with the 
'true' absorption band (dotted line) the apparent band will be a distorted one (Fig. 36); 
a shift in the maximum is very likely. This is known as the Christiansen* effect. It 




1060 cm" 1 
Fig. 36. Christiansen effect for the CH band in iodoform (CHI3) 

may be eliminated or reduced by finely powdering the sample until the particles are 
much smaller than the wavelength of the light being used. 

Instead of grinding the sample and the KBr, other mixing techniques may be 
applied (adding a few drops of a volatile solvent; freeze-drying) to overcome the 
problems of the particle size. 

ORIENTATION EFFECT 

When a crystal sample, or a film in which the molecules are orientated, for instance 
polyethylene, is rotated through a few degrees perpendicular to the radiation beam, 
significant changes in the intensities of the bands may be observed. Some bands 
more or less disappear; others show enhanced intensities. Two factors contribute to 
this phenomenon, called the orientation effect: the instrumental polarisation and the 
orientation of the molecules with respect to the polarised beam. Parallel bands (vibra- 
tions in which the variation of the dipole moment is parallel to the polarised beam) 
will absorb strongly while the opposite holds for perpendicular bands. Hence it is 
clear that rotation of the sample in the beam will cause changes in its spectrum. 

* W. C. Price and K. S. Tetlow, /. Chem. Phys. 16, 1157 (1948). [Christiansen, Wied. Ann. 
(1884).] 



56 SAMPLING 

Only a complete distortion of the orientation of the sample molecules can nullify the 
effect. 

In the mull technique this phenomenon is usually absent. Though all the small 
particles are themselves orientated with respect to the radiation, the overall effect 
will be zero due to the random distribution. Sometimes, however, the particles are 
orientated somewhat as result of mechanical forces and/or gravitation. The effect, 
though small, can then be observed. 

QUANTITATIVE ANALYSIS 

Introduction 

Those expecting infrared spectroscopy to be a powerful method in quantitative 
analysis will be disappointed. Ultraviolet spectroscopy is far more useful in this res- 
pect, but infrared has some important features that make it widely applied. U.v. is 
about a thousand times more sensitive than i.r. but, unlike i.r., is non-specific. Obviously 
the analysis of pure compounds is done by u.v. (provided the sample is u.v. active !) 
whereas mixtures can best be analysed by i.r. Due to the numerous absorption bands 
of the different compounds some bands specific for each component can nearly always 
be found. Despite many problems such as the choice of a solvent, the concentration, 
the solubility, the path length, the background or base line etc., the technique should 
be used as long as no other or better one is available. 

Beer's law 

The theoretical relation between the amount of light from a monochromatic beam 
that passes through an absorbing medium and the amount of absorbant present is 
given by the Lambert-Beer-Bouguer law, often called simply Beer's law: 

1= h exp (-kcl) (1) 

where h is the intensity (or energy) of the radiation incident on the absorbant, / the 
transmitted intensity, c the concentration of the absorbant, / its thickness (path length) 
and k a conversion constant. The formula can be transformed into : 

In Ijh = -kcl or (2a) 

log hll = eel (2b) 

where e = k log e. 

The law is restricted to monochromatic light and for non-interacting absorbants, 
and hence for very dilute solutions. In practice however it turns out that small devia- 
tions from these conditions are possible without serious drawbacks. 

Frequently E (extinction), O.D. (optical density) or A (absorbance, IUPAC nota- 
tion) is used for log hi I, thus leading to the linear relation: 

A = eel (3) 

A plot of A versus c keeping / constant will be a straight line. Substituting c = 1 
mmole/ml and / = 1 cm one obtains 

A = s 

and since A is dimensionless, e will be in cm 2 /mmole; it is often referred to as the 
molar absorption coefficient. It may vary largely from peak to peak but in general it 
lies between and 300 cm 2 /mmole. 



57 
Transmittance and extinction 

Most spectrophotometers produce spectra linear in per cent transmittance (%T). 
Since T = Ijh eqn. (3) can be changed into 

^ = log(100/%r)=ec/ (4) 

e.g. %r = 50 A = 0-30 

%r=io >f = l-oo 

T will vary between 100 and 0%, corresponding to a variation in A from to oo. 
For a background that does not coincide with the 100% transmittance line, a normal 
situation, the calculations seem to be more complex, but in fact are as simple as 
before. 

Suppose the background is found at 7i, the peak maximum at T%. The A for the 
band is A2 corrected for the background, or mathematically 

Attand — A2 ~ A\ (5) 

Substituting A% and A\ one gets 

^band = log (100/%r 2 ) - log (100/%7i) (6a) 

or 

^band = log —=: (6b) 

/o*2 

Let us conclude this section with an example that is often a source of misunderstanding. 
A doubling of the concentration or the path length does not halve the band intensity! 
Consider a band varying from 100 to 50%. As we saw before A = 0-3. A doubling 
of the concentration will lead to A = 0-6. This means that log (100/ %T) = 0-6 or 
that T = 25%. So the band intensity is only increased by 25%! 

Extinction and accuracy 

A further look at the relation between T and A shows that there is a most sensitive 
and accurate range for A as well as very inaccurate ranges. This can be seen in Table 2, 
where A is calculated assuming that T values are accurate within ± 1 % absolute. 

Table 2 



%r±l ,4x103 Percentage 

accuracy 



99 


4-5 ± 4-5 


±100 


90 


46 ± 5 


11 


80 


97 ±6 


6 


70 


155 ± 6 


4 


60 


222 ±7 


3-2 


50 


301 ±9 


30 


40 


398 ± 11 


2-8 


30 


523 ± 15 


2-8 


20 


699 ± 22 


31 


10 


1000 ±46 


4-6 


2 


1699 ± 238 


14 



58 



SAMPLING 



The smallest error is made for A values between 0-4 and 0-5, corresponding to 
40 and 30%T respectively. In general, the range between 60 and 30% transmittance is 
considered to be satisfactory for quantitative analysis. Of course, this type of error is 
not the only one, but we have mentioned it here as it is a specific one, the result of the 
inaccuracy of the instrument and Beer's law. 

Slit width and true absorption band 

Not only the resolution (cf. p. 41) but also the shape of a band is sensitive to the 
slit width. The true absorption band is the curve that one would obtain if it was 
possible to measure the absorption at each wavenumber ai, a%, as . . . a n , i.e. with 



100 



T 

(%) 



f 




<Jb 


J 




!/ 


slit 













Fig. 37. Normal shape of an absorption band 



monochromatic light. In general the band will have a Gaussian or Lorentzian form 
and the absorption for en, <r 2 , as • • • etc. is seen by dividing the curve in infinitely 
small segments of wavenumbers a\, 02, as • • • (see Fig. 37). 

Suppose a slit could be set such that monochromatic light came through (un- 
fortunately it is impossible to do so as diffraction phenomena occur). Varying the 
wavenumber of this beam and measuring the quantity (percentage) of absorbed light 
for each wavenumber will produce the true band. In fact this does mean, however, 
that the 'slit is passing the true band' (Fig. 37). 



59 

For non-monochromatic light passing a slit, i.e. the real situation, the deformation 
of the true band can now be predicted. First, say two frequencies a v and a a pass 
together; the average wavenumber a n = {a v + o q )(2 is indicated on the spectrometer 
or in the spectrum (see Fig. 38). 

For o q = o a the absorption starts, but a n is still greater than a a and for a v = Ob 
the absorption ends, but here a n < <r& and hence the band is broadened. For a n = ffmax 
the absorption will have an average value that will be always smaller than the true one. 




Fig. 38. The band contours for both narrow and broad slits 

In practice, such small deviations from monochromaticity hardly influence the band 
contour. For broader slits (many wavelengths at once), however, the situation is 
worse. The deviation from the true band will be considerable; the apparent peak 
being broad and less intense (see Fig. 38). Hence it is clear that the extinction depends 
to a large extent on the slit width; very unfortunate for quantitative work. From 
practice one knows that as long as the spectral slit width is smaller than one-fifth of 
the half bandwidth of the true band this band will still be produced. 

The spectral slit width Aa i is defined as the difference between the two extreme 
wavenumbers a m and a n at half-height in a plot of the distribution of energy passing 
the slit against the wavenumber with the instrument set at a fixed wavenumber (see 
Fig. 39). 

The half bandwidth of the true band is defined in a similar way. If this band is 

5A 



60 



SAMPLING 



very sharp it will be impossible to get satisfactory reproduction as the slit cannot be 
set so narrow. A distorted band will be the result and quantitative work without a 
standard curve may then be a difficult job. 

It is found, however, that the band area, i.e. the integrated absorbance 



f 

Jo. 



A a da 



is more or less independent on the slit width and is thus a more suitable value for 
quantitative work. Integration can only be done for A — a curves and not for %T — a 




Fig. 39. Energy distribution as a function of the wavenumber for the radiation passing 
the slit at an indicated wavenumber a 

curves. These have first to be transformed into A — a which is a time-consuming 
job. For isolated bands the choice of 01 and 02 is usually simple. 

For quantitative work on different spectrometers the integrated absorption is to be 
preferred. For normal work using just one machine the A ma . K value will do; this saves 
a lot of calculation and measurement. 

Band choice 

A few remarks are necessary concerning the choice of a band upon which the calcula- 
tions will be based. The band should be: 

(a) strong: the stronger it is, the lower the detection limit(s). 

(b) broad : it must not be sharp, as a very small deviation from the wavenumber 
maximum will result in a serious error in A. 



61 

(c) isolated from other bands, otherwise interference wit,b. those bands may occur. 

(d) outside regions where compensated absorption occurs. In addition to solvent 
absorption there is also that due to atmospheric carbon dioxide and water 
vapour. 

(e) free from any effects such as there are in hydrogen bonding, dissociation, etc. 
except if measurements on these phenomena are to be obtained. 

If one or more of these requirements cannot be fulfilled, one will have to be satisfied 
with a less accurate determination of the concentration, though obviously this is 
better than none at all. 

Base line 

For completely isolated bands the background line, i.e. the line obtained under fully 
identical conditions but without a sample, can easily be drawn. Often the peak being 
measured is anything but isolated. A band as represented in Fig. 40 is not rare. In this 




Fig. 40. Probable base lines for an absorption band 



case a base line is used instead of the unknown background line. It is obvious that 
several lines can be chosen in this way. The best fit will be found from practice when 
a A-c plot is prepared. There is no particular reason why one should be better than 
another. Hence there is no need to look for the 'real' base line. As long as the calibra- 
tion line, A versus c, seems to be normal, the choice of the base line is arbitrary. 



62 SAMPLING 

Internal standard 

If for reasons of practice the path length is unknown (mull, KBr technique, uncali- 
brated cells), an internal standard may be added to the sample. This standard should 
only have a few bands and at least one intense band free from any other (e.g. KCN, 
where the C=N band is used). The standard is added to the mull, the KBr or the 
solvent in an exactly known quantity beforehand. The extinction of the unknown 
sample is now determined with respect to the extinction of the standard. The path 
length cancels out this way. For instance when the path length is doubled both A's 
will be doubled, but the ratio will still be the same, indicating that the concentration is 
unchanged. 

Here too a calibration line, A a&mp ie/A standard versus c sam pie, has to be made in 
advance. The method is rather complicated and moreover it is doubtful whether the 
mull and KBr techniques are appropriate at all for quantitative work. 

Cell calibration 

For the calibration of the path length of a cell the interference fringe method is used. 
A parallel monochromatic beam perpendicular to the cell windows is passed through 
the cell. Part of the beam, however, is reflected backwards at the four surfaces of the 
window material. The light passing the rear surface of the first window will interfere 
with the light reflected by the front of the second window. Dependent on the wave- 
length and the path length, there will be destructive interference if 2d = (n + |)A 
where dis the cell thickness in fim, k the wavelength in /nm and n is 1, 2 . . . any integer. 



Anri4 




3500 3000 2500 cm- 1 

14 

Fig. 41. Cell calibration pattern, d = — — - = 001 cm 

2(3300 — 2600) 

Varying A continuously will give rise to the appearance of an interference pattern 
(see Fig. 41), from which d can be calculated using the following formula: 

2(<7l - (7 2 ) 

where <j\ and ai are the wavenumbers between which the number of fringes, An, is 
counted. This method will only give good results provided the windows are flat and 
parallel to each other. Otherwise distorted patterns or no pattern at all result. 



63 

Attenuated total reflection 

When a light beam in a medium of high refractive index is reflected at the interface with 
a medium of a lower index, an evanescent wave is set up in the latter medium propa- 
gating parallel to the interface for a very short distance. Eventually the wave returns 
to the first medium, making the reflection total. This phenomenon can be fully under- 
stood from Maxwell's law for electromagnetic radiation. 

Now suppose that the evanescent wave when propagating in the second medium is 
slightly absorbed. The reflection will then no longer be total but attenuated. This is 




\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\S> 

Fig. 42. Single reflection in an a.t.r. crystal 

called attenuated total reflection (a.t.r.). This attenuated beam can give us information 
on the second medium. 

The use of a.t.r. in infrared spectroscopy started in 1 960 with Fahrenfort's* and 
Harrick'sf papers on theory and practice. The basic idea of the measurement is given 
in Fig. 42. 

A ray or beam perpendicular to medium I is reflected at the interface with medium 
II as result of the difference in refractive index. The angle of incidence 6 is greater 




Fig. 43. Multiple internal reflection 

than the critical angle. The ray comes out again perpendicular to the polished surface. 
Since the penetration depth of the reflected beam (or more correctly, the eva- 
nescent beam) is approximately its wavelength, the effective path length is small 
(a few /um) and thus the absorption will also be small. This can be improved simply 
by increasing the number of reflections. The principle of such methods is given 
in Fig. 43. It is commonly referred to as multiple internal reflection (m.i.r.). The path 
for one ray through medium I is indicated. The number of reflections can be altered by 
changing the angle of incidence. Obviously more sample (medium II) is needed to 
cover medium I on both sides. 

* J. Fahrenfort, Spectrochim. Acta 17, 698 (1961); 18, 1103 (1962). 
t N. J. Harrick, Phys. Rev. Letters, 4, 224 (1960). 



64 SAMPLING 

Several a.t.r. units fitting most commercial spectrometers are now available. 
The principle of m.i.r. has been widely accepted. We will mention briefly the dis- 
advantages, limitations and, more important, the advantages of this technique for i.r. 
work. 

The a.t.r. material (medium I) is a critical component. The desirable characteristics 
are well known, but hard to fulfil. The material should be chemically inert (no 
reaction with the sample), pure (no absorption), quite tough and easy to polish and, 
last but not least, the ratio «materiai/«sampie should be greater than unity, otherwise 
there will be no reflection at all, although it should not be too high, as the attenuation 
is inversely proportional to it. Well-known a.t.r. materials are KRS-5, germanium, 
AgCl or AgBr and silicon. Others may be chosen for special purposes. 

Restrictions as to the sample are less stringent. In general all materials that can 
make good contact with the a.t.r. plate will do. If the peaks are too strong, the sample 
area can be diminished for better results. Powders and rough samples will cause difficul- 
ties, as can be appreciated. 

The only drawback of the a.t.r. technique results from the a.t.r. accessory itself. 
It causes a lengthening of the light path of the sample beam (the double beam character 
becomes lost) and a loss of light. 

On the other hand, a.t.r. has advantages over the transmission techniques, 
especially for opaque materials. The method can be applied to different samples 
such as polymer films, coatings, fabrics, paints, pastes, leather and many others. 
Sampling itself is perhaps even simpler than that for the transmission method. 

The intensity of a band in an a.t.r. spectrum depends on (a) the penetration depth 
into the sample and (b) the ratio of the refractive indices. Both vary with the wave- 
length but the latter can show dramatic changes when 'nearing or passing' an absorp- 
tion band (cf. p. 54). For that reason part of an a.t.r. spectrum may look like a trans- 
mission one (the resemblance may be striking) but the overall appearance will be 
different. 

Quantitative work is hardly possible unless only moderate or low accuracy can be 
accepted. The biggest problem is the reproducibility in mounting the a.t.r. plate and 
the sample, as the effective area of the sample must remain constant. 

Micro attachments are available now, and these sets will make the use of the 
a.t.r. technique still more popular. 

BIBLIOGRAPHY 

R. G. J. Miller and B. C. Stace, Laboratory Methods in Infrared Spectroscopy 2nd edn., 

Heyden & Son, London, 1972. 

L. May, Spectroscopic Tricks, Plenum Press, New York, 1967. 



5 

Interpretation of spectra 



The interpretation of spectra, i.e. correlating spectra with molecular structure, is in 
many ways an art. Fortunately, some general rules can be outlined to facilitate 
learning this art, although some feeling and a lot of experience are required to be able 
to do it really well. 

First let us recall some results from theory. In a spectrum two rather different 
types of absorption bands are present: those attributable to distinct parts of a mole- 
cule ('group frequencies') and those caused by vibrations of the molecule as a whole 
('skeletal modes'). 

The group frequencies are localised at certain regions in the infrared spectrum 
and, as can be appreciated, are very useful indeed for the identification of the functional 
groups of a molecule. The skeletal frequencies are characteristic of a particular 
molecule and as such are not localised at all, though in general they do occur below 
1500 cm -1 . (This region, therefore is often referred to as the 'fingerprint' region.) 

From the preceding it may be clear that the interpretation of a spectrum will 
have to start with the identification of the more or less localised group frequencies. 

A functional group can give rise to none, one, or more than one absorption band, 
depending on the nature of the group. For instance a symmetrical \C=C<^ system 
will not absorb any radiation, the transition being forbidden in the infrared. A G==N 
group leads to one and only one band (the G=N stretching vibration) in the 2200 
cm -1 region, whereas a \CH2 group will give rise to at least 6 bands as we saw 
before (see Chapter 2). 

Let us now apply the above principles, taking the C=C group as an example. 
The band is to be found in the 1650 cm -1 region. 

(a) No C=C band in the spectrum. Three possibilities arise: 

1 . The molecule lacks this functional group. 

2. The group is present but the transition is infrared inactive. 

3. The transition is allowed, but the intensity of the band is too small to be seen 
under the applied conditions. 

(b) One C=C band in the spectrum. Again three possibilities: 

1. Only one C=C group is present in the molecule. 

2. Two or even more fully identical groups form part of the molecule; the bands 
coincide. 

3. As under 1 or 2 but other inactive C=C groups are present. 

65 



66 INTERPRETATION OF SPECTRA 

(c) Two or more C=C bands are found in the spectrum. Two possibilities arise: 

1. At least two optically different types of C=C bands are present in the molecule. 

2. As under 1 but also considering the foregoing possibilities (e.g. inactive groups 
may form part of the molecule). 

The same holds for other groups except that there are many groups without 
forbidden transitions such as C=N, C=0, etc. 

For functional groups of the complex type (CH 2 , CH 3 , N0 2 , NH 2 , etc.) the situa- 
tion is similar. Instead of one band several bands will appear at different regions in the 
spectrum, each band indicating the presence of the group. The bands are not indepen- 
dent and would thus seem to be superfluous or at least useless for interpretation. 
These bands are far from useless, however, as we will see later on. 

To return to the functional group regions, how do we know the region(s) where 
a certain functional group will absorb radiation? This is not as difficult as it looks. 
From the early days of infrared spectrometry data has been compiled on this subject. 
These data are tabulated in numerous tables (see Appendix D), but they can also 
be presented on correlation charts (see Appendix C). On such charts the group frequency 
regions are indicated by lines. Functional groups showing more bands are thus 
represented by more lines. The length of a line denotes the broadness of the region. 

When looking over the correlation charts one can see that many regions overlap 
showing that different functional groups can show absorption bands that coincide. 
Though it is obvious that the identification of a certain band is hampered that way, 
one can make use of the earlier mentioned extra bands. For instance, a CH3 group 
absorbs at 2900 cm -1 and so does a CH2 group. Since a CH3 group will have also a 
band at 1350 cm -1 and a CH 2 group will not, evidence is found for one of the two 
possibilities. Actually the situation is more complicated than this usually, and there 
are usually more than two possibilities. 

So far, only the frequency (wavenumber) of a band has been considered. The in- 
tensities can give new or additional information. A rough idea on the intensity can be 
obtained from this rule-of-thumb : the higher the dipole moment involved in the 
vibration of the functional group, the higher the intensity of the band. Carbonyl 
(C=0) bands are strong, aromatic C — H bands weak, whereas — C=C — bands 
have variable intensity, generally speaking. It will be clear that the presence of more 
identical groups in the same molecule alters the situation. For instance a C=0 
group in one of the high alkanones will show only medium intensity. Furthermore 
the preparation technique has not been taken into account; concentration differences 
as well as scanning conditions can change the intensity of a peak. Obviously the 
use of the band intensity rather than its frequency is a difficult task, not least as a 
result of the arbitrarily chosen terms weak, medium, strong, etc. Experience is an 
important factor in this field. 

A third parameter useful in the interpretation of functional groups is the shape 
of a band. Although the shape depends on the intensity and moreover is highly 
influenced by the representation of the spectrum - linear in wavelength or wave- 
number, scanning conditions - some functional groups can be recognised at once by 
their shapes while the choice is often harder to make in other cases. A few examples 
of some characteristic absorption patterns are given in Appendix B. 

As to the tables and the charts found in the literature and here in this book, some 



67 

preliminary remarks should be made. The tables and thus the charts are obtained 
by simply compiling data from the literature. In general these data are not critically 
reviewed before quoting. Scanning conditions are not taken into consideration at 
all. This is an extremely unsatisfactory situation for the following reasons: 

(a) most data in the literature are given without mentioning the accuracy. 

(b) data are obtained from spectrophotometers with different resolution. 

(c) the spectrophotometers are often insufficiently calibrated. 

(d) data that do not match are compared with each other, using different phases, 
different solvents, etc. 

These and other causes significantly lower the value of the tables. This situation 
will change, but for the time being use must be made of the available data. 

The tables in Appendix D are based on data from the literature as well as from 
the author's own sources. They are presented in a rather condensed form; more 
details can be found in standard works (Appendices A and E). 

We now return to the problem of the unknown spectrum. There are no special 
rules for the beginner and so he will have to think along the same lines as the expert. 
Some guidance here may be very helpful. The most important rule is : think logically. 
Consider all possibilities you can think of, even the rarest ones, and eliminate them 
one by one if evidence is found for this in the spectrum. 

Start with the identification of the functional groups. Do not forget the preliminary 
data, such as the phase, the colour, the smell or its origin, etc. Proceed by drawing up a 
provisional molecule and consider whether this molecule would give a similar spectrum 
or not. If not, start again. If so, try to find a spectrum of this compound in reference 
collections or anywhere in the literature and compare this with that of the unknown 
compound. If there is a difference try to find out the reason for that and change the 
provisional formula. Compare again with a reference spectrum. In case it is not at 
hand try to find the substance and see by personal measurement what the spectrum 
looks like. Proceed until two identical spectra - the unknown and the reference - lie 
on the desk. Only then may the interpretation be called successful. 

BIBLIOGRAPHY 

A. J. Baker and T. Cairns, Spectroscopic Techniques in Organic Chemistry, Heyden & Son, 

London, 1966. 

A. J. Baker, T. Cairns, G. Eglinton and F. J. Preston, More Spectroscopic Problems in 

Organic Chemistry, Heyden & Son, London, 1967. 

L. J. Bellamy, The Infrared Spectra of Complex Molecules, Methuen, London, 1960. 

L. J. Bellamy, Advances in Infrared Group Frequencies, Methuen, London, 1968. 

K. Nakanishi, Infrared Absorption Spectroscopy, Holden-Day, San Francisco, 1962. 



Appendix A 



REFERENCE SPECTRA: MAJOR COLLECTIONS 
Sadtler Standard Spectra 

Sadtler Research Laboratories Inc., have for many years been the leading publishers 
of spectral data. Their collections cover Infrared, Ultraviolet, Nuclear Magnetic 
Resonance and Differential Thermal Analysis. All Sadtler Standard Spectra are 
continuing data projects and all spectra are scanned at SRL. Spectra are available 
printed on paper or in Microfilm as Film or Fiche. 

Two Infrared Collections are available: (1) Infrared PRISM Standard Spectra and 
(2) Infrared GRATING Spectra. The IR Prism collection now totals 41,000 Spectra 
and by annual subscription a further 2,000 spectra are added each year. The IR 
Grating collection was started more recently and apart from an issue of 2,000 spectra 
of new compounds each year an additional 1,000 spectra are issued of compounds 
which were previously scanned for the prism collection. Subscribers can choose 
whether they want the full subscription of 3,000 spectra annually or the non-dupli- 
cating 2,000 spectra. The IR Grating Collection now comprises 22,000 spectra. 

The Sadtler Collections are well supported with excellent indices. These give 
reference to all Sadtler Standard Spectra which may be available in the IR Prism, IR 
Grating, Ultraviolet, Nuclear Magnetic Resonance or Differential Thermal Analysis 
collections. A recent addition is that the 8,000 spectra published by the Coblentz 
Society, printed and distributed by SRL, are also included in these 'Total' indices. 
Retrieval is possible with four different Sadtler Total Indices by: (a) Molecular 
Formula, (b) Chemical Classes (c) Alphabetical Name, (d) Numeric Serial 
Number. 

A most valuable index permitting the identification of unknown spectra is the 
Sadtler 'Spec-Finder'. With the aid of this Spec-Finder searches of spectra can be 
made by the positions of absorption bands in the unknown. 

Smaller Special Collections of Sadtler Infrared Spectra comprise ATR - Atten- 
uated Total Reflectance Spectra; Biochemicals ; Steroids; Inorganics and Organo- 
metallics. 

Whereas the Sadtler Standard Spectra comprise only pure organic compounds, 
the Sadtler Commercial Spectra collections consists of groups of spectra of com- 
mercially available products, most of which are available as Prism or Grating Spectra. 

These separately available groups of commercial spectra are: 

Agricultural Chemicals Drug and Drug Extracts Fibres 

Coating Chemicals Dyes, Pigments & Stains Food Additives 

68 



69 

Intermediates Petroleum Chemicals Rubber Chemicals 

Lubricants Pharmaceuticals Solvents 

Monomers & Polymers Plasticisers Surface Active Agents 

Natural Resins Polyols Textile Chemicals 

Perfumes & Flavours Pyrolysates Water Treatment Chemicals 

Each IR Commercial Spectra group has its own index. Two additional Composite 
indices listing all commercial Spectra from all groups are also available : (a) The Com- 
mercial Spectra Alphabetical and Molecular Formula Index and (b) The Commercial 
Spectra Spec-Finder. 

DMS - Documentation of Molecular Spectroscopy 

The second major collection is the DMS System. The spectrum and relevant data are 
printed on Needle-Sort Cards. Because the System has grown to over 19,000 Spectra 
the collection has become too comprehensive for this method of retrieval. Whereas the 
Sadtler spectra are all run at SRL under standard condtions the DMS spectra are 
collected from different sources, fulfilling, however, certain requirements set by the 
DMS Editorial Board. The chart presentation is smaller than Sadtler but one can 
profit from the listing of wavenumbers (or wavelength) which frequently appear 
alongside the spectrum. Printed lists of indices with supplements are available but 
DMS Index Cards comprising peep-hole cards aim to simplify sorting for reference 
spectra. Each Card has a grid for 5,000 punchable positions. The serial number of 
each spectral card corresponds to one of the positions on the peep-hole card, while 
each peep-hole card corresponds to a property of the code of which there are 211. 
An additional service is the DMS Literature Service which lists all relevant litera- 
ture on infrared, Raman and microwave spectroscopy that has appeared since 1963. 
The DMS System is published by Butterworths and Verlag Chemie. 

API/TRC-The American Petroleum Institute Research Project 44 and 
the Thermodynamics Research Center Data Project 

Both projects issue initial sets of looseleaf sheets in six categories, revised and updated 
semi-annually. Both are located at the Thermodynamics Research Center. 

The American Petroleum Institute Research Projection 44 (API RP 44) compiles, 
calculates, critically evaluates and publishes tables of selected physical and thermo- 
dynamic property values and selected spectral data in 5 categories for the classes of 
hydrocarbons and certain classes of organic nitrogen hydrocarbon derivatives of 
interest to petroleum and petrochemical industries. The collective title of the publica- 
tions is 'Selected Values of Properties of Hydrocarbons and Related Compounds'. 
Infrared spectra fall under Category B; there are 3079 spectra currently available. 

The Thermodynamics Research Center Data Project (TRC), (formerly The Manu- 
facturing Chemists Association Research Project) publishes similar data broken down 
into the same categories as API RP 44 in 'Selected Values of Properties of Chemical 
Compounds' for the classes of organic compounds other than hydrocarbons and for 
certain other classes of inorganic compounds of interest to the chemical industry. 
There are 665 infrared spectra currently available in this collection. 

IRDC-The Infrared Data Committee of Japan 

These cards are also of a needle-sort type. The set now comprises 10,000 cards and a 
further 1,000 spectra are issued per annum. 



70 APPENDIX A 

Coblentz Society Spectra 

This Society collects spectra from various sources and examines and classifies this 
data. More recently, from spectrum 5,001 onwards the spectra are critically evaluated 
to conform to minimum standards of Class III, set by the Coblentz Society. Spectra 
1-5,000 Selected Spectra; 5,001-8,000 Critically Evaluated Spectra. These 8,000 
spectra are printed and distributed for the Coblentz Society by Sadtler Research 
Laboratories and are now also indexed in the SRL Total Indices. The Spectra are 
available on paper or in Microform as Film or Fiche. 

Infrared Spectra of Selected Chemical Compounds — Microform Edition 

A smaller complete collection of about 2,000 spectra recorded linear in wavelength 
(wavenumber tables supplied) of well-chosen simpler type compounds, invaluable for 
laboratories that have problems in establishing a large collection, either due to lack of 
funds or because their work is only in specialised fields. An ideal teaching aid, the 
spectra were formerly available in a paper edition arranged simply in Serial Number 
order. The new low-cost edition is available on Microfilm or Microfiche and it is 
important to note that the spectra have been re-arranged into a Chemical Classes 
Order which permits easy comparison of the spectra of related compounds. An index 
provided also permits the retrieval of spectra by Name, Molecular Formula, Chemical 
Class or Serial No. 

Published by Heyden & Son. 

REFERENCE SPECTRA: MINOR COLLECTIONS 

Bellanato and Hidalgo: Infrared Analysis of Essential Oils contains 214 infrared spectra of 

essential oils which can be compared using a split binding with any of 60 of their constituent 

compound spectra. Heyden & Son, London, 1971. 

Dobriner et al. An Atlas of Steroid Spectra contains 760 spectra, Vol. 1, 1953, Vol. 11, 1958. 

Wiley-Interscience, New York. 

Haslam and Willis : Identification and Analysis of Plastics contains 300 spectra of plastics 

and resins. Iliffe, London, 1965. 

Holubek : Spectra Data and Physical Constants of Alkaloids. A continuing data project. 

Each spectral sheet shows ultraviolet and infrared spectrum. Issues 1-8 cover Sheets 1-1000 

in five binders. Heyden & Son, London. 

Hummel and Scholl: Infrared Analysis of Polymers, Resins contains 1,758 spectra of polymers, 

resins and additives. Wiley-Interscience, New York, 1969. 

Neudert and Ropke: Atlas of Steroid Spectra. Contains 900 spectra of steroids. Springer- 

Verlag, Berlin, 1965. 

Welti: Infrared Vapour Spectra (Group frequency correlations, sample handling and the 

examination of gas chromatographic fractions) contains 300 spectra of volatile organic 

compounds. Heyden & Son, London, 1970. 

JOINT INDICES FOR ALL PUBLISHED INFRARED SPECTRA 

ASTM -American Society for Testing and Materials 

Three universal indices have been issued by ASTM covering 92,000 Spectra available 
in any of the collections or in the original literature. 

AMD-31 Molecular Formula List of Compounds, Names and References to Published 
Infrared Spectra. This is undoubtedly the best buy for anyone who wishes to 
locate spectra in collections or in the literature. The book comprises 616 pages 
and is relatively inexpensive. 



71 

AMD-34 Alphabetical List of Compound Names, Formulae and References to Published 
Infrared Spectra. The first section is an alphabetical list of organic compounds 
with molecular formulae. 

The second section is an alphabetical list of organic compounds, where the 
molecular formulae are unknown. It also includes inorganic compounds in 
alphabetical order. 

The third section again gives cross reference to the original literature in which 
abstracted spectra appeared. 

AMD-32 Serial Number List of Compound Names and References to Published Infrared 
Spectra. It is a companion to the ASTM Spectral Data File, since it identifies a 
compound from the spectrum serial number obtained by searching this data file. 

IRSCOT System 

A different approach to infrared spectral interpretation is used by the Miller and 
Willis IRSCOT System (Hey den & Son). This consists of data cards on each charac- 
teristic group frequency together with a correlation table index, permitting rapid 
access to this data. Each data card contains concise and reliable information on the 
infrared absorption bands of a structural group and gives appropriate literature 
references and examples. Band positions are quoted in both frequency and wave- 
length. The ten sections of the system so far published cover : 



1. Hydrocarbons 

2. Halogen Compounds 

3. Oxygen Compounds (excl. acids) 

4. Carboxylic Acids and Derivatives 

5. Nitrogen (excl. N-O Compounds) 

6. N-O Compounds 



7. Heterocyclics 

8. Sulphurs 

9. Silicon Compounds 

10. Boron Compounds 

1 1 . Phosphorus Compounds (in 

preparation) 



Provision has been made for additional or replacement cards to be issued. The cards 
are supplied in handy binders. 



Publisher'' s addresses* 

American Society for Testing and Materials, 
1916 Race Street, 
Philadelphia, 
Pennsylvania 19103, U.S.A. 

Butterworth & Co., 

88 Kingsway, 

London W.C.2, England. 



Data Distribution Office 

Thermodynamics Research Center, 
Texas A & M Research Foundation, 
F.E. Box 130, 
College Station, 
Texas 77843, U.S.A. 

Sadtler Research Laboratories Inc., 
3316 Spring Garden Street, 
Philadelphia, 
Pa. 19104, U.S.A. 



Heyden & Son Ltd., 
Spectrum House, 
Alderton Crescent, 
London N.W.4, England. 

* Sadtler Standard Spectra, API/TRC Spectra, IRDC Spectra, Coblentz Society Spectra, and the 
ASTM Infrared Indices are distributed in Europe by Heyden & Son, from whom further information can 
be obtained. In all other cases, contact publisher. 



Appendix B 



TYPICAL BAND CONTOURS 




3600 3200 2800 2400 

Wavenumber (cm"') 
Carboxylic acid (Solid in KBr) 



72 



73 



80 



60 



40 



20 



XT 




3600 



3200 



2800 2400 

Wavenumber (cm "') 



Carboxylic acid, long-chain molecule (solid in KBr) 

The broad band is caused by the strongly bridged OH stretching vibration 

of the carboxylic acid group, on which the CH stretching peaks of the 

long chain fatty acid are superimposed. 



74 



APPENDIX B 



80 



60 



40- 



20- 



%T 




3500 3000 2500 2000 

Wavenumber (cm _1 ) 
Amino acid (solid in KBr) 

This very broad band is caused by the strongly bridged OH stretching 
vibration of the carboxylic acid group and the nitrogen-hydrogen stretch- 
ing in the NH 2 and N + H 3 -groups. Besides there may be the peaks of the 
CH stretching band in the normal regions. 



75 



80 



60 



40 



20 



%T 





3600 



3200 



3600 3200 2800 

Wavenumber (cm "' ) 



Left: Free and slightly bonded OH stretching bands at about 3600 and 
3400 cm -1 respectively (solution in CC1 4 ). 

Right: The broad OH stretching band of the bridged OH group in an 
alcohol (pure liquid). 



76 



APPENDIX B 




3000 



2800 



2000 



1800 



800 625 

Wavenumber (cm" ) 

Typical patterns related to mono-substituted benzene rings. 

Left: Above 3000 cm -1 the CH stretching vibrations of the aromatic 

hydrogen atoms. 
Middle: Overtone pattern. 

Right: Out-of-plane bending vibrations of the 5 adjacent hydrogen atoms 
of the aromatic nucleus. 

The patterns as such can be different or may even be absent in many cases 
especially if hetero atoms such as oxygen are present. 



77 



80 



60 



40 



20 



%T 





2000 



1800 



900 



800 



.-H 



Wavenumber I cm" 

Two regions in the spectrum from which para substitution may be 
recognised (pure liquid). 

Left: The overtone pattern, with a dominant peak at about 1900 cm -1 . 
Right: A strong band caused by the out-of-plane bending vibration of the 

two adjacent hydrogen atoms attached to each side of the aromatic 

nucleus. 



78 



APPENDIX B 



100 



80 



60 



40 



20 



%T 





1400 1350 



1400 1350 

Wavenumber(cm" J 



Left: isopropyl or ^em-dimethyl group (pure liquid) 
Right: tert-buty\ group (pure liquid) 

The bands caused by the symmetric CH bending vibration of the CH 3 
groups will appear if the methyl groups are attached to a saturated carbon 
atom. The distance between the split peaks is 16 ± 4 cm -1 for the iso- 
propyl and 27 ± 4 cm -1 for the terf -butyl group. 
The pattern is frequently distorted by other methyl groups in the molecule. 



79 





3000 



2500 



3500 3000 

Wavenumber (cm" 1 ] 



Left: CH stretching region : methylene) groups attached to nitrogen (pure liquid) 

Right: NH stretching region of primary amines (solution in CC1 4 ) 

In compounds such as triethylamine, piperidine, dimethylpentylamine, etc. 

the CH stretching region appears to be extended with peaks at wavenum- 

bers below 2800 cm -1 . 

The NH 2 group has two bands, i.e. asymmetric and symmetric stretching 
bands in the region 3600-3300 cm -1 of different intensity. 



80 



APPENDIX B 




3200 



2800 

Wavenumber (cm" 1 ) 



C — H stretching region 



81 




3000 



2800 

Wa venumber (cm "' ) 



An aldehyde group gives rise to a peak of medium intensity apart from the 

normal CH stretching region at thelow wavenumber side. It appears at 

about 2700 cm -1 . Pure liquid spectrum. 



Appendix C 

CORRELATION CHARTS 



83 



HYDROCARBONS 



-i — i — i — iiii 



MICRONS (u) 



8 9 10 



15 20 3040 



WAVENUMBER (cm~») 4000 3500 3000 2500 20b0 1800 1600 1400 1200 KXM 800 600 400 



Paraffins 



Methyl 
Methylene 
Single C-H 



Alicyclics 



Cyclopropane 
derivatives 



defines 



Linear 

unconjugated 



Vinyl 

Vinylidene 

Vinylene 
trans/cis 

Cyclo 

Conjugated 

Allenes 
Acetylenes 
Aromatics 

Polynuclear 



— I 1 1 1 1 1 i i i i i i ' 

WAVENUMBER (cm-') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 



MICRONS (u) 



8 9 10 



15 20 3040 



84 



APPENDIX C 



HALOGEN COMPOUNDS 



MICRONS (u.) 


I 1 1 1 1 i r — i — i 1 1 1 — i — i — 1 i — i — ~i 

7 8 9 10 11 12 13 14 15 20 25 30 3540 5b 70 100 


(cm-') 




WAVENUMBER 


1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 

i i i i i i i i t i t i t i 


Fluorine 




C-F bonds 




sat. 






unsat. 










. 1 ■--! 


cyclo 






polymeric 






aromatic 




Chlorine 




C-CI bonds 




sat. 






unsat. 




► ' 


cyclo 


i • *q-. 




polymeric 




aromatic 




Bromine 




C- Br bonds 




aliphatic 
aromatic 




^ 


.1 ' 


r ( 


• > 


Iodine 


, — 


di 




Acid halides 




fluorine/chlorine 

chlorine 

bromine 


' 


, 












WAVENUMBER 


i i.i ■ * i i i i i ii ii 

1400 1300 1200 1100 1000 900 800 700 600 500 400 300 200 100 


(cm" 1 ) 




MICRONS (u.) 


7 8 9 10 11 12 13 14 15 20 25 30 354050 70100 

.i .... i. .. i r i i i i i i i i • 1 i i i 



85 



OXYGEN COMPOUNDS 

(excluding carboxylic acids) 



MICRONS (\i) 



WAVENUMBERCcm - ') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 
i l i I l I I I 1 1 1 I 1 — 



Alcohols Primary 

Secondary 
Tertiary 

Phenols 

Ethers Non-cyclic 

Cyclic 

Epoxides 

Cyclic 
dioxy cmpds. 

Acetals 



Peroxy Hydroperoxides 
cmpds. 

Peroxides 



Ozonides 

Aldehydes Sat. 
Unsat. 
Aromatic 

Ketones Sat. 
Unsat. 
Aromatic 
rf-Halogen 

Sterically strained 
cyclopropyl 



Cyclic 



Ketenes 



ot&B-Diketones 
(keto form) 

oi-Hydroxy aromatic 
carbonyl cmpds. 

oi.,B-Unsat. amino 
ketones 

Quinones 



free 



H -bonded 



213-48crrr 



WAVENUMBER (cm-") 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 



MICRONS (n) 



9 10 15 20 3040 

_i i i i i i i ' ■ ' 



86 APPENDIX C 

CARBOXYLIC ACIDS AND DERIVATIVES 



MICRONS (|i) 



WAVENUMBER (cm-') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 

— I 1 1 1 i 1 1 1 I I i i i 



Carboxylic acids 

Mono- 
Di- 
Per- 
Salts 

Anhydrides 

Acoyl & aroyl peroxides 

Acid halides 

Chloroformates 

Carboxylic acid 
esters 

Alkyl 

Alkyl conj 

Aryl conj 

a. -Hydroxy 

Lactones 



free ass. 



free ass. 



„ — i 1 1 1 1 1 1 1 1 i 1 1 1 — 

WAVENUMBER (cm-') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 



MICRONS (}!) 



8 9 10 



15 20 3040 
j_i i i i_ 



87 



NITROGEN COMPOUNDS 

(excluding NO compounds) 



MICRONS ((« 



15 20 3040 



WAVENUMBER(cm') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 

i i i i i i I I I I I 1 1 — 



Amines Primary 

Secondary 
Tertiary 

Amine salts 

Unsat. amino ketones 

Ketimines, 

Azomethines 

Azines, Benzarnidines 

Hydrazines 

Carbodiimides 

Isocyanates 

Azo compounds 

Hydrazo ketones 

Thioamides & salts 

Azides 

Nitriles 

Amides Primary 

Secondary 
Tertiary 

Carbamates 
Ureas 

Polypeptides 

ionised form 
Lactams 

Diacylamines 

Amido acids 

Amino acids 

zwitterion form 
Amino acidsN+salts 



WAVENUMBER(cm') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 



MICRONS (|i) 



6 7 8 9 10 15 20 30 40 

— i 1 1 1 i i i i i i ' ■ 



88 



APPENDIX C 



N-O COMPOUNDS 



MICRONS «i) 



8 9 10 



WAVENUMBER (cm-') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 
i i i i I I I I I I I I I 



Oximes 



Nitroso compounds 



Dimeric nitroso cmpds. 



Azoxy compounds 



Nitrite esters 



Nitrosamines 



Nitrosamides 



Nitro compounds 



Nitrate esters 



Nitrate] 

> salts 
Nitrite 



Nitramines 



Carbonitrates 



— i 1 1 1 1 1 1 1 1 1 1 i i 

WAVENUMBER (cm-') 4000 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 



MICRONS ((i) 



7 8 9 10 
_i i i i_ 



15 20 30 40 



89 



SULPHUR COMPOUNDS 



MICRONS (ji) 


3 


T 

4 


1 

5 


1 1 1 1 — r — 1 — 1 1 1 1 r 1 r- 

6 7 8 9 10 15 30 40 


WAVENUMBER (cm 


4000 3500 3000 2500 2000 

1 1 1 1 t 


1800 1600 1400 1200 1000 800 600 400 

1 I 1 I 1 1 1 1 


Mercaptans 




" 






Thiophenols 




" 






Thioacids Mono 




* 




■ ■ 


Di- 




— 




■ ■ • 


Sulphides.Disulphides 








^r^ 


Thionitrites 
Thiocyanates 




— 




1-1 




Isothiocyanates 




— 






Sulphonyl chlorides 








, . , — . 


Sulphonic Anhyd. 
acids 

Hyd. 

Other- SO2- cmpds. 

Thiol compounds 




n-^n-. 




^^ . . 






Xanthates 








,— .„ 


Sulphinic acids 




r-l r-i 




r-, 


OtherS =0 cmpds. 










Sulphoxides 








- 


Dithioesters 








■ 


Thionesters 








■ 


Trithiocarbonates 








- 


Thioacid halides 
N-C=S compounds 








■-■ 




WAVENUMBER (cm ■ 


1 1 1 1 1 
) 4000 3500 3000 2500 20 


00 


1 1 1 1 1 1 1 1 

1800 1600 1400 1200 1000 800 600 400 


MICRONS (u.) 


3 
1 


4 f 

1 1 




6 7 8 9 10 15 20 3040 
1 1 1 1 1 1 i'i' 1 ■ 1 . 



90 



APPENDIX C 



SILICON COMPOUNDS 



MICRONS (p) 
WAVENUMBER (cm" 1 



1 7 8 9 »b ' ' "is 2b sb'sbio 



100 



) 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200 
J I I | I I I 1 I I 1 1 1 — 



Alkylsilanes 

Silacycloalkanes 

Silyl olefins 

Arylsilanes 

Silanols 

Oxysilanes 

Alkoxychlorosilanes 

Silyl esters 

Siloxanes 

Polysiloxanes 
Metal siloxanes 

Silicates 

Fluorosilicates 

Silyl halides 
Fluorine 
Chlorine 
Bromine 
Iodine 

Silylamines 

Silazanes 

Cyclosilazanes 



Silyl azides& 

isocyanates 

Si-P compounds 



Silyl sulphides 
Silicon thiols 



„ — i 1 1 1 1 1 1 1 1 1 1 1 n — 

) 3500 3000 2500 2000 1800 1600 1400 1200 1000 800 600 400 200 



4 5 6 7 8 9 10 15 20 30 50100 
i i i ■ iii i i i i I i I i_i i 



WAVENUMBER (cm 
MICRONS (y) 



Appendix D 



INFRARED ABSORPTION FREQUENCIES OF FUNCTIONAL GROUPS* 



Functional group 



Absorption range Example 
(cm -1 ) (cm -1 ) 



Example compound 



ALKANES 








a. linear 








CH3 asymmetric 


2970-2950 


2967 


n-Octane 


CH3 symmetric 


2885-2865 


2868 




CH3 asymmetric 


1465-1440 


1466 




CH3 symmetric 


1380-1370 


1380 




CH2 asymmetric 


2930-2915 


2920 


n-Octane 


CH2 symmetric 


2860-2840 


2854 




CH 2 


1480-1450 


1470 




(CH 2 ) W n > 4 


723-720 


723 


n-Octane 


« = 3 


735-725 


733 


n-Pentane 


n = 2 


755-735 


741 


2-Methylpentane 


n= 1 


800-770 


781 


n-Propane 


b. branched 








For CH2, CH3 wavenumbers, 


see section a. 






CH 


2890 


2890 


Triphenylmethane 




1340 


1341 




CH 3 — CH— 


1385-1380 


1384 


2-Methylheptane 


1 


1372-1366 


1366 




CH 3 


1175-1165 

1160-1140 

922-917 






CH 3 


1395-1380 


1393 


2,2-Dimethylhexane 


1 


1375-1365 


1366 





CH 3 — c— 

I 

CH3 
— c — c— 

1 i 

CH 3 CH 3 



1252-1245 

1225-1195 

930-925 

1165-1150 
1130-1120 
1080-1065 



1160 
1122 
1071 



3,4-Dimethylhexane 



* Abbreviations: sp = sharp, br = broad, (w) = weak, (s) = strong. 

91 



92 






APPENDIX D 


Functional group 


Absorption range 
(cm- 1 ) 


Example 
(cm- 1 ) 


Example compound 


CH 3 

1 


1391-1381 
1220-1190 


1389 
1192 


3 , 3 -Dimethy lhexane 


R— C— R 

i 


1195-1185 


1189 




CH 3 








C 2 H 5 


1250 


1250 


3-Ethylhexane 


1 


1150 


1155 




R— CH 

1 


1130 


1131 




1 
C2H5 








c— c— c 

1 


1160-1150 






CH3 








c. cyclic compounds 








Cyclopropane 
derivatives 


3100-3072 
3033-2995 


3075 
3028 


Cyclopropane 




1030-1000 


1024 




Cyclobutane 
derivatives 


3000-2975 

2924-2874 

1000-960 

or 


2974 
2896 


Cyclobutane 




930-890 


901 




Cyclopentane 
derivatives 


2959-2952 
2870-2853 


2951 
2871 


Cyclopentane 




1000-960 


968 






930-890 


894 




Cyclohexane 
derivatives 


1055-1000 
1015-950 


1038 
1014 


Cyclohexane 



For larger rings see section b. 

UNSATURATED COMPOUNDS 
a. isolated — C=C — bonds 

CH 2 =CH— 



CH 2 =C 



\ 



/ 



— CH=C 



\ 



3095-3075 


3096 


3030-2990 


2994 


1648-1638 


1645 


1420-1410 


1420 


1000-980 


994 


915-905 


912 


3095-3075 


3096 


1660-1640 


1661 


1420-1410 


1420 


895-885 


887 


3040-3010 


3037 


1680-1665 


1675 


1350-1340 


1351 


840-805 


812 



1-Butene 



Methylpropene 



3-Methyl-2-pentene 



93 



Functional group 



Absorption range Example 
(cm -1 ) (cm -1 ) 



Example compound 



\ / 




3040-3010 


3030 


cis-2-Butene 


c=c 


(pis) 


1660-1640 


1661 




/ \ 




1420-1395 


1406 




H H 




730-675 


675 




H 




3040-3010 


3021 


trans-2-Butene 


\ / 




1700-1670 


1701 




o=c 


(trans) 


1310-1295 


1302 




/ \ 




980-960 


964 




H 











b. conjugated — C=C — bonds 

— C=C— C=C— 1629-1590 1592 

1820-1790 1821 



1,3-Butadiene 



c. Allenic — C=C — bonds 

— C=C=C— 1960-1940 

1070-1060 



d. — C=C — bonds 






— feC— 


2270-2250 


2268 


e. — C=CH groups 






CH (stretch) 


3320-3300* 


3320 


— C^C — 


2140-2100 


2122 


CH (bend) 


700-600 




AROMATIC COMPOUNDS 




a. general 






CH 


3060-3010 




CH substitution bands, 






overtones 


2000-1650 (w) 




o=c 


1620-1590 sp 
1590-1560 sp 




CH 


1510-1480 sp 
1450 sp 




b. mono-substitution 








1175-1125 


1170 




1110-1070 


1088 




1070-1000 


1032 




765-725 


728 (s) 




720-690 


693 (s) 



2-Pentyne 



1-Butyne 



Toluene 



* CC1 4 solutions only. 
7 



94 






APPENDIX D 


Functional group 


Absorption range 
(cm- 1 ) 


Example 
(cm- 1 ) 


Example compound 


c. di-substitution 








ortho 


1225-1175 
1125-1090 
1070-1000 

765-735 


1185 
1121 
1053 
741 (s) 


o-Xylene 


meta 


1175-1125 

1110-1070 

1070-1000 

900-770 

710-690 


1171 

1095 

1039 
769 (s) 
690 (s) 


ra-Xylene 


para 


1225-1175 

1125-1090 

1070-1000 

855-790 


1219 
1120 
1043 
796 (s) 


/^-Xylene 


d. tri-substitution 








1,2,3- 


1175-1125 

1110-1070 

1000-960 

800-755 

740-695 


1162 
1095 
1009 

765 (s) 
710 (s) 


1 ,2,3-Trimethylbenzene 


1,2,4- 


1225-1175 

1130-1090 

1000-960 

900-865 

855-800 


1156 
1130 
1000 

873 (s) 
805 (s) 


1 ,2,4-Trimethylbenzene 


1,3,5- 


1175-1125 

1070-1000 

860-810 

705-685 


1165 

1039 
836 (s) 
690 (s) 


1 ,3,5-Trimethylbenzene 


e. tetra-substitution 








1,2,4,5- 


870-855 


870 


1,2,4,5-Tetramethyl- 
benzene 


ALCOHOLS 








a. general 








OH unbridged group 

OH inter- and intra- mole- 

cularly H-bonded 
OH intermolecularly 

H-bonded 


3650-3590 sp 
3570-3450 
3400-3200 br 







b. primary alcohols 



1350-1260 
1065-1020 



1339 
1028 



1-Pentanol 











95 


Functional group 




Absorption range 
(cm- 1 ) 


Example 
(cm- 1 ) 


Example compound 


c. secondary alcohols 


1370-1260 
1120-1080 


1369 
1111 


2-Pentanol 


d. tertiary alcohols 


1410-1310 
1170-1120 


1379 
1124 


2-Methylbutanol-2 


e. aromatic ring 


hydroxy compounds 






OH unbridged 
OH dimer 
OH polymer 




3617-3599 sp 
3460-3322 br 
3370-3322 br 
1410-1310 
1225-1175 


1350 
1225 


Phenol 


PEROXIDES 










a. aliphatic 




1820-1810 

1800-1780 

890-820 






b. aromatic 




1805-1780 
1785-1755 
1020-980 






ETHERS 










a. aliphatic 










O— CH 3 
C— O— C 
O— (CH 2 ) 4 
O— CH 3 




2830-2815 
1150-1060 

742-734 
1455 


1140 


Diethyl ether 


b. aromatic 










=c— o— c 
c— o— c 




1275-1200 
1075-1020 


1247 
1038 


Anisol 


c. cyclic 










C— o— C 




1140-1070 






d. epoxides 

trans compounds 
cis compounds 




1260-1240 
890 
830 


1261 
826 


1 ;2-Epoxybutane 
1 ;2-Epoxybutane 



96 



APPENDIX D 



Functional group 



Absorption range Example 
(cm -1 ) (cm -1 ) 



Example compound 



e. tetrahydrofuran derivatives 



1098-1075 
915-913 


1076 
912 


Tetrahydrofuran 


f. trioxans 

1175 
958 


1172 
957 


Trioxan 


g. tetrahydropyran derivatives 






1120-1080 

1100-900 

825-805 







h. dioxan derivatives 



1125 



1122 



Dioxan 



KETALS, ACETALS 








R O— C 

\ / 
C 

/ \ 
R O— C 




1190-1158 
1143-1124 
1098-1063 






KETONES 










a. aliphatic 




1725-1705 

1325-1215 

1200 


1727 
1269 
1215 


Butanone 


b. unsaturated 










C=C 

c=o 




1650-1620 
1685-1665 


1618 
1684 


Methyl vinyl ket 


c. aromatic 










Aryl, alkyl 
Aryl, aryl 




1700-1680 
1670-1660 


1694 


Acetophenone 


d. cyclic 










4- and 5-membered 
6- and 7-membered 


rings 
rings 


1775-1740 
1725-1700 


1739 
1703 


Cyclopentanone 
Cycloheptanone 


e. diketones 










a-Diketones 
/?-Diketones 
y-Diketones 




1730-1710 
1640-1540 
1725-1705 


1721 


Diacyl ketone 









97 


Functional group 


Absorption range 


Example 


Example compound 




(cm" 1 ) 


(cm- 1 ) 




f . halogen substituted 








<x,a-Dihalogen substituted 


1765-1745 






a-Halogen substituted 


1745-1725 






ALDEHYDES 








a. general 










2900-2700 (2 bands) 




CH 


2720-2700 
975-780 






b. aliphatic 








c=o 


1740-1720 


1735 


Butyraldehyde 


CH 


1440-1325 


1390 




c. unsaturated 








0=0 


1650-1620 


1637 


Crotonaldehyde 


C=0 a, /3 unsaturated 


1690-1650 






d. aromatic 








CH 


2750-2720 sp 


2725 


Benzaldehyde 


c=o 


1715-1695 


1701 






1415-1350 


1391 






1320-1260 


1312 






1230-1160 


1203 




CARBOXYLIC ACIDS 








a. general 








OH 


3200-2500 br 






CH 


1440-1396 
1320-1210 






OH dimer 


950-900 br 






C=0 halogen substituted 


1740-1720 






C=0 aliphatic 


1720-1700 


1718 


Acetic acid 


C=0 unsaturated 


1710-1690 


1698 


Crotonic acid 


C=0 aromatic 


1700-1680 


1695 


Benzoic acid 


C=C unsaturated 


1660-1620 


1655 


Crotonic acid 


b. carboxylic ions 








0=0 


1610-1560 






0=0 


1420-1300 






KETENES 










2155-2140 


2155 


Ketene 




1135-1120 


1136 





98 



APPENDIX D 



Functional group 



Absorption range Example 
(cm -1 ) (cm -1 ) 



Example compound 



ESTERS 

0=0 unsaturated, aryl 1800-1770 

0=C unsaturated, aryl 1730-1710 

C— O acrylates, fumarates 1300-1200 

C— O 1190-1130 
0=0 electronegatively 

substituted 1770-1745 

0=0 a, y keto 1755-1740 

0=0 saturated 1750-1735 

C=0 /? keto 1660-1640 

C— O benzoates, 1310-1250 

phthalates 1150-1100 

C— O acetates 1250-1230 

1060-1000 
C — O phenolic acetates 1205 

C— O formate 1 200-1 1 80 



1718 

1282 
1192 



1744 



1277 
1108 
1246 
1047 

1190 



Ethyl acrylate 



Methyl acetate 

Methyl benzoate 
Propyl acetate 

Propyl formate 



LACTONES 






/S-Lactones 


1840-1800 




y-Lactones 


1780-1760 


1776 


<5-Lactones 


1750-1730 






1280-1150 


1168 


ANHYDRIDES 






a. aliphatic 






0=0 


1850-1800 


1842 


0=0 


1785-1760 


1783 


c— o 


1170-1050 


1134 


b. aromatic 






0=0 


1880-1840 


1866 


0=0 


1790-1770 


1773 


c— o 


1300-1200 


1267 


c. cyclic 






0=0 


1870-1820 


1818 


0=0 


1800-1750 


1772 


ACID CHLORIDES 






a. aliphatic 






0=0 


1815-1770 


1802 



Butyrolactone 



Acetic acid anhydride 



Phthalic acid anhydride 



Glutaric acid anhydride 



Acetyl chloride 



Functional group 


Absorption range 


Example 


99 

Example compound 




(cm- 1 ) 


(cm- 1 ) 




b. aromatic 








c=o 


1775-1760 


1773 


Benzoyl chloride 


0=0 


1730-1700 


1726 




AMIDES 








a. primary 








NH free 


3500 






NH free 


3400 






NH bridged 


3350 


3346 


Butyramide 


NH bridged 


3190 


3191 




c=o 


1660-1640 


1660 






1430-1400 


1430 




b. secondary 








NH free trans 


3460-3400 






NH free cis 


3440-3420 






NH bridged trans 


3320-3270 


3280 


iV-Methylacetamide 


NH bridged cis 


3180-3140 






bridged cis, trans 


3100-3070 


3090 




c=o 


1680-1630 


1652 




NH 


1570-1510 


1564 






720 br 


725 




c. tertiary 








0=0 


1670-1630 


1670 


Af,iV-Dimethylformamide 


AMINO ACIDS 








NH 


3130-3030 br 








2760-2530 (not always present) 






2140-2080 






c=o 


1720-1680 






Ionised form 


1600-1560 
1300 






C=0 a-amino acids 


1754-1720 






C=0 /?, y-amino acids 


1730-1700 






Amino acid hydrochlorides 


3030-2500 (more 


bands) 




NH amino acid hydro- 








chlorides 


1660-1590 






NH amino acid hydro- 








chlorides 


1550-1490 






AMINES 








a. general 








N— CH 3 


2820-2730 






N— CH 3 


1426 






C— N 


1410 







100 



APPENDIX D 



Functional group 



Absorption range Example Example compound 

(cm -1 ) (cm -1 ) 



b. aliphatic. 


primary 






NH free 




3500-3200 


3350 






(2 bands) 


3210 


NH 




1650-1590 
1200-1150 


1630 






1120-1030 


1100 



Ethylamine 



c. aliphatic, secondary 



NHfree 

NH 

C— N 
C— N 


3500-3200 
(1 band) 
1650-1550 
1200-1120 
1150-1080 


3230 

1126 
1090 


Dipropylamine 


d. aliphatic, tertiary 








C— N 

C— N 


1230-1130 
1130-1030 


1175 
1070 


Ethyldimethylai 


e. aromatic, primary 


3510-3450 
3420-3380 
1630-1600 


3460 
3413 
1621 


Aniline 



f . aromatic, secondary 

Free 3450-3430 

Bridged 3400-3300 



3400 



iV-Methylaniline 



UNSATURATED NITROGEN COMPOUNDS 
a. imines 

NH 3400-3300 

C=N 1690-1640 



b. oximes 

Liquid 

Solid 

Solid 



3602-3590 
3250 
3115 



Aliphatic 
Aromatic 



1680-1665 

1650-1620 
1300 
900 









101 


Functional group 


Absorption range 
(cm- 1 ) 


Example 
(cm- 1 ) 


Example compound 


CYANIDES, ISOCYANIDES 






C=N unconjugated 


2265-2240 


2256 


Ethyl cyanide 


G==N conjugated or 
aromatic 


2240-2220 


2222 


Benzyl cyanide 


G=N cyanide, thiocyanide 
complex 2200-2000 






N=C alkyl isocyanide 
N=C aryl isocyanide 


2183-2150 
2140-2080 


2166 
2100 


Methyl isocyanide 
Phenyl isocyanide 


CYCLIC NITROGEN 


COMPOUNDS 






a. pyridines, quinol 


ines 






CH 

C=C, C=N 


3100-3000 
1615-1590 
1585-1550 


3030 
1590 


Pyridine 




1520-1465 


1490 






1440-1410 








920-690 


707 






(substituent dependent) 






b. pyrimidines 








CH 


3060-3010 






C=C, C=N 


1580-1520 






Ring 


1000-900 







UNSATURATED NITROGEN-NITROGEN COMPOUNDS 

Azo compounds 1630-1575 



N=N azides 
N=N azides 


2160-2120 
1340-1180 


2130 
1297 


Phenylazide 


NITRO COMPOUNDS 








a. aliphatic 


1570-1500 

1385-1365 

880 


1546 

1362 

879 


2-Nitrobutane 


b. aromatic 










1550-1510 

1370-1330 

849 


1527 

1351 

853 


Nitrobenzene 



102 



APPENDIX D 



Functional 



group 



Absorption range 
(cm- 1 ) 



Example 
(cm- 1 ) 



Example compound 



PHOSPHORUS COMPOUNDS 



O — H phosphoric acids 


2700-2560 br 


P— H 


2440-2350 sp 


p=o 


1350-1250 


p=o 


1250-1150 


P— O— C 


1240-1190 


P— O— R 


1190 


p— o— c 


1170-1150 


P— O— C 


1050-990 


P— O— P 


970-940 


P— F 


885 


P=S 


840-600 


O— P— H 


865-840 


O— P— O 


590-520 


O— P— O 


460^40 



PHOSPHORUS-CARBON COMPOUNDS 



P — C aromatic 


1450-1435 




P— C aliphatic 
P— C 


1320-1280 
750-650 


1298 
707 


PO4 3- aryl phosphates 
PO4 3- alkyl phosphates 
PC>4 3 ~ alkyl phosphates 


1080-1040 

1180-1150 

1080 





Trimethylphosphine 



DEUTERATED COMPOUNDS 

O— D deuterated alcohols 2650-2400 



O — D deuterated carboxylic 
acids 



675 



SULPHUR COMPOUNDS 



c=s 


1400-1300 


1357 


Dithioacetic acid 


s=s 


1200-1050 






P=S 


840-600 






SH mercaptans 
C — S mercaptans 


2600-2550 
700-600 


2580 
665 


Ethyl mercaptan 


C— S— C dialkyl sulphides 


750-600 
710-570 


726 
676 


Methyl ethyl sulphide 




660-630 


654 




Aliphatic sulphones 


1410-1390 
1350-1300 


1407 
1316 


Dimethylsulphone 


Sulphonic acids 


1210-1150 

1060-1030 

650 






S— CH 3 


1325 







103 



Functional group 



Absorption range Example 
(cm -1 ) (cm -1 ) 



Example compound 



SILICON COMPOUNDS 








SiH alkylsilanes 


2300-2100 


2175 


Dimethylsilane 


Si(CH 3 ) 2 


1265-1258 

814-800 

800 


1262 




Si(CH 3 ) 3 


1260-1240 


1259 


Methoxytrimethylsilane 




850-830 


844 






760 


763 




Si — C aromatic 


1429 
1130-1090 






Si— C 


860-715 






Si — O siloxanes 


1100-1000 






Si — O — C open-chain 


1090-1020 






Si — O — Si open-chain 


1097 






Si— O— Si cyclic 


1080-1010 







HALOGEN COMPOUNDS 

a. iodine compounds 

b. bromine compounds 



ca. 500 



700-500 



c. chlorine compounds 

Monochloro 

Fully chlorinated 
compounds 



800-600 
750-700 

780-710 



d. fluorine compounds 



1400-1000 
1100-1000 



Fully fluorinated compounds 745-730 



INORGANIC 


COMPOUNDS 






a. sulphates 


1200-1140 

1130-1080 

680-610 


1143 

1117 

617 


Potassium sulpha 


b. nitrates 


1380-1350 
840-815 


1370 
825 


Potassium nitrate 


c. nitrites 


840-800 
750 







104 



APPENDIX D 



Functional group 



Absorption range Example 
(cm -1 ) (cm -1 ) 



Example compound 



d. water of crystallisation 

1630-1615 



e. halogen-oxygen salts 



Chlorates 980-930 


978 


Potassium chlorate 


930-910 


932 




Bromates 810-790 


793 


Potassium bromate 


Iodates 785-730 


756 


Potassium iodate 


f. carbonates 






1450-1410 


1410 


Calcium carbonate 


880-860 


875 




g. selenium compounds 






Selenates 895 






420 






Selenites 740 






460 







Appendix E 

GENERAL BIBLIOGRAPHY 



BOOKS 

H. C. Allen and P. C. Cross, Molecular Vib-Rotors, John Wiley & Sons, New York, 1963, 
320 pp. 

N. L. Alpert, W. E. Keiser and H. A. Szymanski, Theory and Practice of Infrared Spectroscopy, 
2nd edn., Plenum Press, New York (Heyden & Son, London) 1970, 380 pp. 

C. N. Ban well, Fundamentals of Molecular Spectroscopy, McGraw-Hill, New York, 1966, 
282 pp. 

G. M. Barrow, The Structure of Molecules, Benjamin, New York, 1963, 156 pp. 
G. M. Barrow, Introduction to Molecular Spectroscopy, McGraw-Hill, New York, 1962, 
332 pp. 

N. B. Colthup, L. H. Daly and S. E. Wiberley, Introduction to Infrared and Raman Spectro- 
scopy, Academic Press, New York, 1964, 511 pp. 
R. T. Conley, Infrared Spectroscopy, Allyn & Bacon, Boston, 1966, 293 pp. 

B. W. Cook and K. Jones, A Programmed Introduction to Infrared Spectroscopy, Heyden & 
Son, London, 1972. 

A. D. Cross, An Introduction to Practical Infrared Spectroscopy, 3rd edn., Butterworths, 
London, 1969, 110 pp. 

J. R. Ferraro, Low Frequency Vibrations of Inorganic and Coordination Compounds, Plenum 
Press, New York (Heyden & Son, London) 1971, 309 pp. 

Ian Flemming and D. H. Williams, Spectroscopic Methods in Organic Chemistry, McGraw- 
Hill, London, 1966, 215 pp. 

L. A. Gribov, Intensity Theory for Infrared Spectra of Polyatomic Molecules, Plenum, 
New York, 1964, 120 pp. 
M. L. Hair, Infrared Spectroscopy in Surface Chemistry, Dekker, New York, 1967, 314 pp. 

D. N. Kendall, Applied Infrared Spectroscopy, Reinhold, London, 1966, 560 pp. 

K. E. Lawson, Infrared Absorption of Inorganic Substances, Reinhold, London, 1961, 227 pp. 
L. H. Little, Infrared Spectra of Absorbed Species, Academic Press, London, 1966, 428 pp. 

C. E. Meloan, Elementary Infrared Spectroscopy, McMillan, London, 1963, 180 pp. 

K. Nakamoto, Infrared Spectra of Inorganic and Coordination Compounds, John Wiley & 

Sons, New York, 1963, 330 pp. 

W. J. Potts Jr, Chemical Infrared Spectroscopy, Vol. 1 : Techniques, John Wiley & Sons, 

New York, 1963, 322 pp. 

H. Siebert, Anwendung der Schwingungsspektroskopie in der Anorganischen Chemie, Springer 

Verlag, Berlin, 1966, 209 pp. 

W. W. Wendlandt and H. G. Hecht, Reflectance Spectroscopy, Interscience, New York, 

1966, 275 pp. 

R. G. White, Handbook of Industrial Infrared Analysis, Plenum Press, New York, 1964, 

452 pp. 

R. Zbinden, Infrared Spectroscopy of High Polymers, Academic Press, New York, 1964, 

260 pp. 

105 



Index 



Absorbance, 56 
Absorption coefficient, 56 
Absorption, infrared radiation, 3 

integrated, 60 

process, rotational energy, 13 
Accidental degeneracy, 23 
Amplifier, 38 
Amplitude of a mode, 1 1 
Anharmonicity constant, 1 1 
Anharmonic oscillator, 1 1 
Anomalous dispersion, 3 
A.P.I, collection, 69 
A.S.T.M. collection, 70 
Atmospheric absorption, 43, 45 
Attenuated total reflection, a.t.r., 63 
Automatic gain control, a.g.c, 45 
Automatic slit control, a.s.c, 45 
Automatic speed suppression, a.s.s., 39 

Balance control, 39, 43 

Band choice, 60 

Band contour, 59 

Base line, 61 

Beer's law, 56 

Bending vibration, 26 

Black body radiation, 33 

Blaze angle, 36 

Boltzmann's distribution law, 10, 14 

Born-Oppenheimer approximation, 17 

B values, 7 

Calibration, cell, 62 

transmission, 46 

wavelength, 46 

wavenumber, 46 
Cavity cell, 49 
Cell calibration, 62 
Chopper, 38 
Christiansen effect, 54 
Coblentz Society Spectra, 70 
Collections of spectra, 68 



Comb, 38 

Combination band, 23 
Conformation, 53 
Conversion factors, viii 
Correlation charts, 82/J 

Dead spot test, 43 
Degeneracy, 23 
Degrees of freedom, 6 
Detector, 36 
Diatomic molecules, 6 
Dipole moment, 16 

change in rotation, 12 

change in vibration, 4, 16 
Dispersion, 3 

grating, 35 

prism, 35 
Dissipation of absorbed radiation, 5 
D.M.S. collection, 69 



Echelette grating, 36 
Electromagnetic energy, 1 

radiation, 1 

radiation, interaction with matter, 3 

spectrum, 2 

wave, 1 
Energy, anharmonic oscillator, 1 1 

dissipation, 5 

harmonic oscillator, 7 

kinetic, 7 

limited, 44 

photon, 1 

potential, 7 

rotational, 11, etc. 

units, viii 

vibrational, 9, 1 1 

vibrational-rotational, 1 7 
Entrance slit, 33 
Excited state, 9 
Exit slit, 33 
Extinction, 56 



106 



107 



Film technique, 50 
Filter, 35, 45 
Force constant, 7 
Frequency, 1 
Frequency, units, viii 

vibrational, 7 
Functional group, 25 
Fundamental vibration, 6 
Fundamentals, number of, 6 

Gain, 42 
Gas cell, 48 
Gas phase, 19 
Gas techniques, 48 
Gaussian band, 58 
Globar, 31 
Golay detector, 38 
Grating, 35 

blaze angle, 36 

condition for diffraction, 35 

constant, 35 

echelette, 36 

order, 35 
Ground state, 9 
Group frequency, 25 

Half bandwidth, 60 
Harmonic oscillator, 7 

energy levels, 9 

selection rules, 9 
High resolution, 44 
Hooke's law, 7 
100% line, 39 
100% setting, 43 
Hydrogen bonding, 51 

intermolecular, 52 

intramolecular, 52 

Index of refraction, 3 
Intensity, /, 56 

integrated, 60 

rotational band, 14 

vibrational band, 16 
Interference fringes, 62 
Intermolecular bridge, 52 
Internal standard, 62 
Interpretation of spectra, 65 
Interpretation tables, 91 
Intramolecular bridge, 52 
I.R.D.C. cards, 69 
Irscot system, 70 
Isomerism, 53 
Isotope effect, 23 
Isotopic, shift, 24 

/, rotational quantum number, 12 
Johnson noise, 37 



Kinetic energy, 8 

Large molecules, 26 

Limited energy spectra, 44 

Linear molecule, number of fundamentals, 6 

Liquid cell, 49 

phase, 21 

techniques, 49 
Lorentzian band, 58 
Luffs formula, 42 

Mass, relation to frequency, 7 

Mecke collection, 70 

Micro cell, 49 

Mull technique, 50 

Minimum volume cell, 48, 49 

Mirrors, 33 

Molecular interaction, 19 

vibration, 18 
diatomic, 18 
polyatomic, 24 
Moment of inertia, 12 
Monochromator, 33 

lay-out, 32 
Multi-pass cell, 48 

n, vibrational quantum number, 9 

Nernst glower, 31 

Nichrome coil, 3 1 

Noise, 38, 42 

Noise equivalent power, n.e.p., 38 

Non-rigid rotator, 14 

Normal vibration, 6 

Nujol mull technique, 50 

Optical density, 56 

lay-out double beam spectrometer, 32 
Optimum transmission for quantitative mea- 
surements, 57 
Order of diffraction, 35 
Ordinate expansion, 45 
Orientation effect, 55 
Oscillation, 42 
Overshoot, 42 

method, 43 
Overtones, 11 

P branch, 18 

Paper shrink and stretch, 47 

Parallel band, 55 

Perpendicular band, 55 

Photon, 1, 13 

Planck's constant, viii 

curve, 33 

formula, 31 
Polarised radiation, 55 



108 

Polyatomic molecules, 24 

Polyethylene, 29 

Polymers, 30 

Polymorphism, 50 

Polystyrene, 28 

Potassium bromide technique, 50 

Potential energy, 7 
curve, anharmonic oscillator, 1 1 
curve, harmonic oscillator, 8 

Preamplifier, 38 

Preparation techniques, 48 

Pressure effect on gas spectra, 21 

Prism, 35 

Protein spectrum, 29 



Q branch, 18 
Quantitative analysis, 56 
Quantum of radiation, 1, 13 

Radiation, black body, 33 

electromagnetic, 1 

energy, 1 

frequency, 2 

interaction with matter, 4, 13 

scattered, 44 
R branch, 18 
Recording paper, 47 
Reduced mass, 7 
Reference beam attenuator, 44 
Reflection of materials, 34 
Refraction, 3 
Refractive index, 3 
Resolution, 39 

effect on band shape, 41 
Resolving power, 39 
Response time, 38 
Rigid rotator, 1 1 
Rocking vibration, 26 
Rotation, 6, 1 1 
Rotational constant, 12 

energy, 12 

energy levels, 13 

quantum number, 12 

vibrational bands, 18 

vibrational energy levels, 18 

Sadtler collection, 68 
Sample preparation, 48 

gas, 48 

liquid, 49 

solid, 49 

solution, 49 

vapour, 48 
Scale expansion, 45 



INDEX 



Scanning conditions, 39 

speed, 39 

time, 43 
Scattered light, 44 
Scissoring vibration, 26 
Selection rules, 6 
Servo motor, 38 
Servo system, 41 
Signal-to-noise ratio, 42 
Skeletal vibration, 25 
Slit, 33 
Slit programme, 35, 39 

width and resolution, 41 

width and true band, 58 
Solid phase, 21 

technique, 49 
Solution technique, 49 
Solvent, 49 
Source, 31 

Spectra, collections of, 68 
Spectral slit width, 59 
Stefan-Boltzmann law, 32 
Steroid spectrum, 27 
Stray light, 44 
Stretching vibration, 26 
Structural isomerism, 53 
Symmetry, 22 

Term scheme, 12, 15 
Thermocouple, 37 
Torsion vibration, 26 
Translation, 6 
Transmittance, 57 
Transmission, %T, 57 

of materials, infrared, 34 
Triatomic molecules, 22 
True absorption band, 58 
Twisting vibration, 26 
Types of vibration, 26 
Typical spectra, 72 

Units, viii 

Velocity of light, 1 
Vibrational frequency, 7 

quantum number, 9 

types, 26 
Vibration, 6 
Vibration, anharmonic, 11 

diatomic molecules, 6 

energy levels, 9, 10 

harmonic, 7 

Wagging vibration, 26 



109 



Wavelength, 1 Window material, 34 

Wave nature of light, 1 

Wavenumber, viii % control, 39 

Wien's law, 32 Zero-point energy, E , 9 



Further Heyden Publications 
on Infrared Spectroscopy 



I RSCOT SYSTEM 

by R. G. J. Miller and H. A. Willis 

Currently covering ten chemical classes, the popular and widely renowned I RSCOT 
system is a quick and simple-to-use reference guide consisting of correlation tables, a 
master index and data cards providing telegram-style information on specific infrared 
bands. 

INFRARED VAPOUR SPECTRA 
by D. Welti 

This important work shows why vapour spectra are as readily applicable to structural 
determination as liquid or solid spectra. The author's lucid description of relevant tech- 
niques and instrumentation is accompanied by over 300 fully indexed vapour spectra. 

INFRARED ANALYSIS OF ESSENTIAL OILS 
by J. Bellanato and A. Hidalgo 

This is a study of the application of infrared spectroscopy to the characterisation of 
essential oils and their constituents. It covers 35 essential oils, and the text is illustrated 
by over 200 spectra of essential oils and their constituent compounds. 

LABORATORY METHODS IN INFRARED SPECTROSCOPY (2ND EDN.) 
by R. G. J. Miller and B. C. Stace 

First published in 1965, this book rapidly established itself as a standard reference work 
for all who use infrared spectroscopy. In this revised and greatly enlarged second edition, 
eminent spectroscopists acquaint you with a wealth of short cuts and 'tricks of the trade' 
that have saved them hours of needless experimentation. 

A PROGRAMMED INTRODUCTION TO INFRARED SPECTROSCOPY 
by B. W. Cook and K. Jones 

In an exciting and logical manner, this unique publication instructs students and tech- 
nicians in the theory and practice of infrared spectroscopy. The reader, working at his 
own pace, tackles the subject step-by-step - absorbing each concept before progressing to 
the next - until he is ultimately able to prepare a sample, set-up and operate a spectro- 
photometer and interpret the resulting spectrogram. 



For more details on publications on infrared spectroscopy write to : 
Heyden & Son Ltd., Spectrum House, Alderton Crescent, London NW4 3XX 
Heyden & Son GmbH., Steinfurter Strasse 45, 4440 Rheine/Westf., Germany 



About the author: 



J, H. van der Maas studied chemistry at the Utrecht State University, and in 
1959 gained his degree. He remained at the University to lecture and carry 
out research, and he received his doctorate in 1965. He is flow in control of 
the research and instruction on molecular spectroscopy in ihe University's 
Department of Analytical Chemistry. 



Heyden St Son Ltd., Spectrum House, Alderton Crescent, London NW4 3XX 
Hoyden & Son GmbH, Steinfurter Strasse 45, 4440 Rheine/Westf., Germany 



03 

i 
ft 



About this book: 1 



■c 
i 

n 



Bask Infrared Spectroscopy was first published in 1969 to provide an authori- 
tative guide for those just starting infrared spectroscopy, and a convenient 

source of correlation data for students. This second edition was produced as a 5 

direct result of the reaction of readers to the original book, and it incorpo- 
rates many of their helpful suggestions, Basically, the main texi is unchanged 
and covers fundamental theory, instrumentation, sampling and spectral inter- 
pretation. The appendices devoted to reference spectra, correlations and 
charts are augmented by an enlarged appendix on typical band contours (with 
explanatory text) and an entirely new general bibliography of relevant litera- 
ture, 

'It will serve as a good introduction to infrared spectroscopy." 

Laboratorv Practice